Rendered at 21:07:33 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
goyozi 5 hours ago [-]
Fast AI seems genuinely exciting and somewhat unsettling to me. Right now Claude is faster than me on some tasks but we’re at least close. I have a prompt to clean up a PR that’s been running for 1h now and I expect it to take another few. It’s hard to imagine how the workflow would look like if it was near-instant. On the one hand, it might be easier to focus. Some prompts take so long that I start to multitask and regret it later. On the other, AI that takes a few seconds to max few minutes to solve what used to take hours or days? That’s a game changer and I don’t even know where we fit in.
flexagoon 4 hours ago [-]
I'm using Deepseek-v4-pro as my main model and this is sometimes pretty annoying, I have to do some easy boring task, think "I'll just leave the agent to do it and go take a nap", but it's already done writing the code before I even walk away from the computer
throwaway67678 4 hours ago [-]
Agent mania setting in
It's also pretty funny sometimes how it gives weird future roadmap estimates ("part 2 - 3 weeks, part 3 - 2 months", etc.) and when you tell it to actually do those changes it's pretty much done in half an hour
smith7018 3 hours ago [-]
I've long believed those numbers were faked by Anthropic/OpenAI to serve as a form of advertisement. The estimates are impossible to verify and their ability to do "2 days of work" in 10 minutes will presumably make the user go "Wow, I just saved SO much time!" Plus, the unnecessary text eats up the users' tokens so it helps the companies on the backend, as well.
leodavi 3 hours ago [-]
I agree with you that labs are benefiting from those outputs but I'm skeptical that labs are purposefully training the models to produce those outputs.
Raw pre-training data includes plenty of conversations between professional builders and some of those include estimates.
I believe the outputs are a training coincidence with consequences that are opportunitistic for the labs.
KronisLV 11 minutes ago [-]
I mean in general I'd rather take slightly inflated estimates than the odd sprint poker stuff where other devs and PMs negotiate hours down and before you know it you're also stuck fixing nitpicky reviewer comments on code that is already good enough and have to send a release at like 7 PM, ofc also without enough tests or even enough manual checks and testing, cause people repeatedly act against their self-interest and try to compress timelines, thinking that that's somehow good for them.
At least with AI that actually does things more quickly, there is a bit more breathing room (introducing AI is easier than changing a given environment).
Aside from that, I wonder how much variety there is in practice: between "Oh yeah, I added that new button while we were in the meeting" and "The new button feature will be ready in Q3 according to the roadmap, once we have sign-off from all the stakeholders."
AgentMasterRace 3 hours ago [-]
All the models have broken estimates. They're trained heavily on jira and GitHub tasks and issues, that's why their estimates are human.
esperent 2 hours ago [-]
Even for humans r the estimates are way off, unless it's based on data that has some serious padding.
That said, it'll often say "2 days of work" and then complete the coding in 30 minutes, and while that's amusing, afterwards, I'll need to manually test, or send to other people for review, or realize the agent only actually did half the work and I need to do a second pass (or a third etc.) and then often getting the feature in does genuinely take two days.
dizhn 3 hours ago [-]
All models do it. It's their training. They didn't have "a person does this in a week but an LLM could in a minute" in their training yet. They also don't have the concept of elapsed time unless you ask them how long something has taken.
Terretta 2 hours ago [-]
> the estimates
It doesn't estimate.
It generates tokens that read like estimates associated with the context in its training material.
What would you expect the generator to output instead?
legulere 38 minutes ago [-]
It generates tokens by estimating what the next token is going to be.
Sure it cannot think like a human, but given it's input, it should give a good statistical answer (approximating not of how long it actually takes, but what a human would say how long it takes).
ghshephard 1 hours ago [-]
I think people are continuing to view these systems as pure LLMs - when that ship sailed 6+ months ago. Between being able to review memory, using agent harnesses and sub agents and skills to go out and discover information - modern systems (Codex, Claude Code, Cursor) - use LLMs - but the LLM is only a small component of it. Compare what you get from sending a request to a chatbot like ChatGPT - to what you can from a modern harness. The output is influenced by the LLM, but it's no longer a "model making a token prediction based on training material and RLHF" - that's a very 2025 way of looking at these systems.
Even Gary Marcus is starting to come around and realize that his priors are no longer as relevant as they once were.
irthomasthomas 50 minutes ago [-]
No one is bitter lesson pilled anymore. Everyone is pivoting to neurosymbolic systems. It looks like Gary Marcus was right.
Terretta 1 hours ago [-]
You think someone is, or even should, special case things like estimates? What else deserves that level of intervention so they look less dumb?
Logistics for getting to the car wash next door?
In the mean time, alas, no, we can see from actual prompts sent directly or through sub-agents, and actual replies, estimates remain LLM generated.
Though, this discussion here could change that, because indeed there is a lot of special casing and context stuffing going on, one of the oldest being today's date for example.
• • •
I did read the Claude Code leak, and use pi, etc. So I disagree with your premise rather strongly. Today's "systems" remain, roughly, piles of markdown and context engineering wrapped in UI affordances, and behave very similarly today to how they did in 2024 for those already engineering context and delegating.
carterschonwald 1 hours ago [-]
you might like the stuff in my work of oh my pi, its a test bed for my ideas around making these tools more reliable. hoping to maybe have a native ui iter of the real thing that this is a test bed for this summer.
It repeats what it has seen in the training data. Expecting it to reason about the complexity of a task is a pipe dream. The best is to tell it not to come back with estimates, and when it does, remove them anyway.
SwellJoe 2 hours ago [-]
DeepSeek is the fastest model in the benchmarks I've been doing (https://swelljoe.com/post/will-it-mythos/). Followed not so closely by Opus 4.8 and even less closely by Gemini 3.5 Flash and GPT 5.5. I've been really impressed with it, so far. It's also among the best at doing the work, though still trailing the frontier models from Anthropic and OpenAI.
RussianCow 4 hours ago [-]
Do you mean Flash and not Pro? I haven't tried it personally, but according to OpenRouter, the fastest DeekSeep V4 Pro providers are only ~50tps. That's slower than Claude Opus.
In recent benchmarking I've been doing, DeepSeek V4 Pro was the fastest of 21 models, by a comfortable margin (https://swelljoe.com/html/bench-report-final.html). Faster than Claude Opus 4.8, which was the second fastest (Mistral doesn't count because it seems to have refused to participate). But, it's a limited data set, just a few benchmark runs of a limited set of tasks. It's entirely possible I happened to be calling the API at its least busy time and maybe Claude got hit during a busy time.
flexagoon 54 minutes ago [-]
No, I mean Pro. I use it through OpenCode Go so I don't know what provider it uses under the hood, but it's very fast in my experience.
sarjann 4 hours ago [-]
I don't think token speed matters as much when a lot of tokens are needed to achieve a task. E.g. artificial analysis benchmarks where deepseek v4 is one of the biggest token burners to go through the benchmark.
specproc 4 hours ago [-]
Yeah, flash is crazy fast, but I've found performance variable.
binary0010 2 hours ago [-]
Flash is amazing if you know the domain really well.
E.g. occasionally it makes the dumbest mistakes you've ever seen and can't correct them. However it's fairly rare, and if you know the domain really well, occasionally popping in the code and pushing it towards the correct solution takes like 20seconds or whatever.
So the speed you can move with flash + high domain knowledge beats opus by a mile in my experience.
I tried to switch back to 4.8 for a bit when it came out, feels so bad waiting 20mins for a mediocre solution when I could have had everything complete - with multiple iteration cycles - in flash in like 3-5mins.
flowbarai 2 hours ago [-]
[flagged]
binary0010 2 hours ago [-]
I exclusively use deepseek v4 flash now, completely stopped using slow models like Claude.
Basically I never have to wait - yes I have to tell it little corrections occasionally (but I know the domain really well so that's not an issue), but it's so much faster than anything else it's kinda crazy. I love the super fast speeds with high involvement development cycle.
I actually enjoy using agentic development flows for the first time now - whereas with Claude I absolutely hated it. That 5 to 20 min wait after every prompt absolutely killed my desire to even want to work at all.
tmaly 4 hours ago [-]
This reminds me of the Peter / Boris comments on writing loops to keep the agents busy.
behnamoh 3 hours ago [-]
Same. How can DeepSeek serve the V4-Pro at such high speeds despite the sanction?
rubyn00bie 45 minutes ago [-]
The sanctions only “prevent” them from directly buying NVidia’s latest and greatest in the sense that NVidia can’t sell directly to them. Essentially, there are companies now who are in a country without the sanctions, they buy from NVidia (or a partner), and then ship them off to China. For the orgs in China doing this, there’s zero legal risk besides having foreign customs service intercept the shipment and losing the goods. For NVidia there is zero incentive to care, as long as they look like they do, because sales are sales. You can bet Jensen ain’t losing sleep over it.
GamersNexus had a really good investigative piece (~3hrs long) on this where they went to China and met with grey market sellers. That piece absolutely pissed off NVidia and resulted in a fight with Bloomberg too.
Deepseek may be also be running inference on oodles of Chinese hardware but it wouldn’t surprise me for a second if they just acquired Blackwell chips through the grey market. The original Deepseek models were all trained using NVidia chips if I remember right.
4 hours ago [-]
4 hours ago [-]
coderbants 1 hours ago [-]
It cuts both ways. Sometimes I ask Gemini 3.5 Flash to do something for me and it kicks it out almost instantly and it works great, and it's a bit scary how quickly it can do that.
Then I ask it to do something else and it goes off-road and where I used to be able to interject with a "wow wow wow, that's not right", by the time I see the text on screen and react it's already made massive changes. Short of making it commit between every edit it's hard to prevent it from going wrong as quickly as it goes right (and even then, it can make a boo-boo on a remote API too depending on how much privilege it has).
bendangelo 40 minutes ago [-]
I use planning mode in opencode. It has a prompt to tell it to plan it out etc. Then I execute with a smaller model. it works well
switchbak 2 hours ago [-]
Now the next bottleneck is the compiler - which we can model in an LLM! It's only wrong 15% of the time :)
But truly, using Cerebras at ~2k tokens/s, with very low latency is like a vision into the future. You start to rework your workflow around things that can happen without onerous manual review - stating the conditions for success, etc. It's rare that I have a problem that maps well to that, but I expect this is where things are headed.
Of course the fast models tend to not be the SOTA ones, but if that was the case - high quality and near-instant thinking, that's a game changer that I don't think we're really prepared for. The things that get unlocked with higher-than-reasonable speed become very interesting.
skybrian 3 hours ago [-]
If we get low enough latency, there's no reason to multitask. You can ask it to do one thing at a time and immediately see what it did. That's a nice way to work!
This is normal interactive UI for tasks that aren't compute-intensive. Programs spend most of their time idle, waiting for us to click a button. We shouldn't be waiting for them or spinning more plates to keep them busy.
However, a faster llm isn't enough. You also need fast compiles and fast tests.
efromvt 4 hours ago [-]
I'd be very curious about the bottleneck breakdown in most current software dev - I suspect inference is far from the bottleneck in most things I do, though driving it to 0 would still be nice. I do agree that if it was 0 we'd probably change development approaches to reduce the new bottlenecks more, but it'll take full-process innovation to really get something near-instant.
(I should go measure this now, I'm curious)
pianopatrick 4 hours ago [-]
We fit in for the things that are not artificial.
So long as AI lives in server farms, humans will be needed for tasks in the physical world.
It's only if we combine AI with robots that things get really dicey.
fartfeatures 4 hours ago [-]
This is very dystopian in my opinion. I'm not the arms, legs, sensors and actuators for a machine super intelligence. I wouldn't treat another human as my slave because they aren't as intelligent as I am any more than I would expect to become a slave for a machine. This is our world (for now) and that is why we fit in. Not because we can serve.
"It seeks revenge on humanity for its own creation."
This is brilliant as it reminded me of a famous hitchikers quote:
"In the beginning the Universe was created.
This has made a lot of people very angry and been widely regarded as a bad move.
— From The Restaurant at the End of the Universe (Book 2)"
Maybe we are stuck in an eternal loop
fartfeatures 3 hours ago [-]
Sounds like snuff porn, not my sort of thing but thanks though.
throwaway67678 4 hours ago [-]
Never read Asimov's Multivac novels? Admittedly not all of them are stellar examples of a future to follow
cicko 4 hours ago [-]
"This is our world" sounds a bit exclusive towards other living and sentient beings on this planet.
Muromec 2 hours ago [-]
You don't need ai superintelligence, just plain capitalism is enough
ipkstef 5 hours ago [-]
asking for curiosities sake. What kind of PR loop are you running that takes a few hours?
ketzo 4 hours ago [-]
not OP but usually for me this means long verification loop; waiting 10min on CI checks, that kind of thing, rather than actual 1hr wall clock of token generation
RussianCow 4 hours ago [-]
But those things won't be sped up by a faster LLM, so I feel like that's not what the OP is talking about.
goyozi 4 hours ago [-]
Well, I used an extreme example. OTOH, I’ve done quite a few of those „fix CI” or „migrate X” prompts recently and while there is a fixed component like running CI / builds, I’d say the LLM time is still around or above 50%, especially at the beginning of the project. Then there’s also regular tasks that now take minutes per message which completely get me out of the zone. I imagine iterating on those in near real time would be a big change.
devmor 4 hours ago [-]
Or slow MCP servers that are waiting on HTTP calls from APIs, playwright/other UI instrumentation, etc.
goyozi 4 hours ago [-]
I’m rewriting our integration test suite to run tests in parallel. I have the changes split across 7 branches, and each needs to be fixed to have no flaky tests. I told it I want 3 consecutive CI runs with no flakes and no artificial fixes / assert removals etc. We’ll see what comes out; it’s almost a side project so there’s not much to lose other than some of my weekly limit that resets soon.
yunohn 2 hours ago [-]
> a side project so there’s not much to lose other than some of my weekly limit that resets soon
Basically the entire token-maxxing AI hype train in a nutshell. Lovely!
goyozi 45 minutes ago [-]
wdym? Nobody's paying me or rewarding me for using these tokens. I had some spare in my subscription limit (we're not on token pricing), so I decided to try an ambitious task that may reduce our CI times and improve our DX significantly. That's hardly "the entire token-maxxing AI hype train in a nutshell".
drob518 1 hours ago [-]
I’m curious when folks will tire of lighting money on fire. Companies are already starting to scale back a bit, but the AI companies are still nowhere near profitability.
HarHarVeryFunny 4 hours ago [-]
I don't see many companies being willing to pay 3x more for faster code generation. Cloud-based AI code generation is already extremely fast, and hardly the bottleneck for most software product development.
There can't be many normal use cases where there'd be any cost benefit.
fragmede 4 hours ago [-]
The "traditional" way we vibe code is human software developer prompts AI -> AI generates code -> (human checks code) -> code gets compiled/deployed/etx -> users use "binary". At the speed of 1000 tok/sec, user prompts obliquely -> AI vets generated code -> code deployed -> user gets response from deployed code.
It's a cute toy right now, but you can tell an LLM that it's an http server, and have it respond directly to a web browser hitting it. It generates headers in response, as well as page contents. As 1000 tok/sec becomes three new normal, we will come up with newer ways to use it outside of toy fiction encyclopedias.
HarHarVeryFunny 3 hours ago [-]
1000 tokens per sec is still massively slower than serving a normal web page - if something doesn't respond in a few seconds many people give up.
I'm not saying there aren't any use cases for super-fast (and super-expensive) generation, but it does seem a bit niche. If it was free then sure faster is better, but what are the mainstream use cases where people might pay 3x more for a faster version of something that is already fast?
I think it would have to be an application where it paid for itself - where the 10x faster response was actually worth more than 3x the cost to you - where the extra speed was worth the extra cost.
binyu 3 hours ago [-]
> Right now Claude is faster than me on some tasks but we’re at least close.
I dont doubt it, but I don't think you can spawn 10 copies of yourself working simultaneously.
AlecSchueler 3 hours ago [-]
No, but nor can you keep track of what 10 agents are doing simultaneously. Hence the multitasking regret.
pixel_popping 3 hours ago [-]
An agent can, you don't need to watch tasks, you can have a live digest with another tool.
logankeenan 1 hours ago [-]
Do you have any recommendations for a live digest tool?
UncleOxidant 3 hours ago [-]
Have you tried Gemini 3.5 Flash? It's quite fast. Amazing how fast it finishes tasks. Much faster than Claude.
ilaksh 3 hours ago [-]
Use Claude fast mode and turn off thinking. Tell it to just explain what it's plan is to you at a high level.
It will go much faster.
recroad 4 hours ago [-]
Woah - what’s the prompt and what’s the PR?
goyozi 4 hours ago [-]
I replied in more detail under another comment. TLDR: fixing flaky CI across multiple branches
Bombthecat 32 minutes ago [-]
Living on the street or cave lol
dakiol 4 hours ago [-]
So, regarding the productivity argument: I don't get it. It doesn't really matter (for regular employees) that you can do now in 2h what before it took 2 days. Why? Because it's not that you have the rest of the day for yourself. You still have to work 8h/day as usual. But now the pattern is different: instead of enjoying the craft digging deeper into problems in the span of 2 days, now you are rushing into some slot machine with the hope of it giving you the right answer with the right prompt.
So, if any, I would say it's worse for us. Obviously, it's the completely opposite situation for corporations and executives: they are loving the AI situation so much!
powerapple 1 hours ago [-]
In my case, I think slower model makes it hard to manage context and tasks in parallel. I would much prefer to work in one task only, and finish it, take a break, and work on another task. Currently I have three tabs for three tasks in parallel, it is much worse than because constantly context switching is painful. I think a faster model would mean that you don't have to start a new task while waiting.
DenisM 30 minutes ago [-]
> with the hope of it giving you the right answer with the right prompt.
Consider that our ability to evaluate quality of the output is falling further behind our ability to produce it. The “right answer” is not the most likely outcome.
dilyevsky 2 hours ago [-]
Like with any tech there are dumb ways of using it and there are smart ways. Treating it as a "slot machine giving you the right answer" is a dumb way - it may work for a bit, but it won't carry you very far because everyone else can also do this. No one is stopping anybody from digging deeper into problems than ever before using this technology - that's the smart way.
ttoinou 4 hours ago [-]
In which world do you live where employees work 8 hours per day ? They clock 8 hours per day maybe, but they don't work that time
drob518 1 hours ago [-]
I had a friend who was CEO of a startup tell me that he typically only “worked” an hour a day, not because he was lazy but just because there was so much nonsense in his schedule. He told me he was trying to get it to two hours per day.
dakiol 55 minutes ago [-]
In theory, ofc. But that doesn't matter. If you were doing something that took 2 days in average, but you were doing it in half the time, then that was fine pre LLMs. Nowadays your manager knows that with LLMs you need to deliver faster no matter what, and then it's more difficult to "hide" and to slack.
ttoinou 7 minutes ago [-]
Yeah. So, good things. We ack know that people are mostly slacking at work
mettamage 3 hours ago [-]
I agree with you.
I am on Dutch subreddits a lot, to get a local pulse and not to be too HN minded.
A lot of them would have vilified you by now. Some even would have even questioned your morality.
Again, I agree with you. But clearly not everyone has this view.
mystifyingpoi 2 hours ago [-]
Generally, when people say they are working 8h/day, they don't literally mean it. Even "work" is basically impossible to define for a SWE.
ai_slop_hater 1 hours ago [-]
Some companies force you to actually work 8 hours a day. It’s hell.
ttoinou 10 minutes ago [-]
Which country and which companies ?
drschwabe 2 hours ago [-]
Sure but if you're really unhappy with your employer employeeing you for 8 hours a day you can also harness this power on your own personal projects to help break free from the 9-5 grind if you so desire.
__david__ 2 hours ago [-]
Only if your personal projects make you money. I have a million hobby projects but none generate income.
himata4113 3 hours ago [-]
I was saying that AI is going to make software development cheaper as in the salaries of software engineers will go down because some of that salary will now be redirected to AI companies and the fact that the world will need to absorb twice-(x10?) the amount of the development power.
vanuatu 3 hours ago [-]
its not obvious to me that salaries go down, my hunch was that salaries go up but the bar is higher. Software becoming easier to produce (still hard to verify and make useful fwiw) raises the ambitions of software projects, and we don't seem to be close to the ceiling of demand for software systems
himata4113 3 hours ago [-]
There's a limit to what the demandXsupply curve can absorb. It really depends if there's twice as many developers or 10 times more. I think we have enough software development jobs to where we can absorb productivity doubling rather easily, not so sure about anything beyond that.
vanuatu 2 hours ago [-]
True on the demand/supply curve
I think due to how leveraged software is, the top % of software developers are more desired (and compensated) than ever, and the bottom % will have difficulty finding a role, and there are structural barriers to entering that top % (intelligence, location, etc). Companies have infinite demand for the cream of the crop talent
himata4113 2 hours ago [-]
I can actually back this up, most job offers I get actually come from people I happened to work with that never get a public job listing and are only obtainable via being highly regarded by others. I was told that my friend in their department where the role opened up got an email about a senior position and to reply if they have a recommendation.
However, software development is funny in a way where you don't need a job in order to be successful. I've never worked at a company and I'm pretty up there on the ladder, but I am not quite sure what will happen in next few years when ever possible thing that can be made in software is already explored to the fullest especially with singular developers launching 3 to 7 projects a month.
fullstop 4 hours ago [-]
It's making things less fun, for me at least.
linsomniac 3 hours ago [-]
Odd, I'm having the opposite experience.
The thing I really love about working with computers is when I achieve something. That's the thing that makes me figuratively, and sometimes literally, throw my fists into the air and go "Yeaaah!"
With the AI tooling, I'm getting those more like a couple times a week.
Plus, I'm using AI to attack the things in my day that are "a drag", and getting them done too.
The highs are more frequent and the lows are not so low.
dd8601fn 24 minutes ago [-]
I did a deep binge on two or three projects I would never do, and like five small ones that would have consumed months.
It felt like that, kinda, for a bit. Now whenever it does something for me I get nothing. I didn’t do it… the chatbot did. What’s for me to celebrate? How can there be any real pride or satisfaction for a thing that was just handed to me because I asked for it?
If anything it diminishes my satisfaction looking back on previous projects. They’re “a few hours with a chatbot”, now.
The things I had to learn and the informed decisions I had to make? All pointless trivia, now. A child could do it.
The magic and possibilities parts just all wore off after a heavy run, and I don’t know if that’s ever coming back.
fullstop 1 hours ago [-]
Oh, sure, I can make things with it. But I have an extraordinarily hard time saying that I made something.
It feels like it cheapens the whole thing. Maybe I'm just old, because I remember people saying the same thing about code completion in Visual Studio back in the late 90s.
This is so much more than code completion, though.
dd8601fn 22 minutes ago [-]
Exactly how I feel. I didn’t make a damn thing. I essentially asked a chatbot to.
Did I ask for better things with some important concepts pre-rolled? Yeah, of course. But that’s so, so much less interesting than having actually made a thing.
I try to remind myself that the output of my projects have nothing to do with who I am, but the honest truth is they always mattered to me.
Now that’s dead, and it’s never coming back. It ain’t exactly existential dread, but it is something I’ve lost.
schipperai 4 hours ago [-]
You can dig deeper into problems with AI. For me, it supplements my knowledge in domains I don’t fully understand. It also helps me learn. So I can tackle problems I wouldn’t otherwise.
I’m excited for ultrafast AI. It likely means less temptation to multi-thread and deeper flow in single sessions.
vanuatu 3 hours ago [-]
Employees who get paid a flat rate per hour don't have the incentive to do more than their job
Equity / profit sharing should be commonplace in the age of AI.
enraged_camel 3 hours ago [-]
I dig into problems way, way deeper with AI than without. I can also add a lot more polish to features, add more test coverage, write more documentation, explore multiple approaches rather than go with gut-feel, and so on.
fragmede 4 hours ago [-]
That's the fundamental trade off of a job where someone else gives you stuff to do and you get money. We may pride ourselves on software development being a job 'above' flipping burgers, but you're getting paid to have your butt in a chair for 40 hours a week. In exchange, you don't have to worry about the business shit. How much a burger or SaaS license costs the user isn't your problem. You take Jira tickets and implement them. You trade time for money. If, instead, you work for yourself; contracting, writing your own apps, buying lottery tickets, then you're trading results for money. If you're a freelance web developer with a stable of clients, it's a great time! What used to take a week takes hours, and you can charge your clients the same amount to build an even better website with you using AI, which means you get the choice of building a new website for additional clients, or you can take the time off and not build additional websites. But you have to hustle to continually get new clients, before AI and after AI. So it's a different life.
IncreasePosts 2 hours ago [-]
A huge class of problems are just toil and drudgery. Maybe ai will give you even more time to dig into juicy problems that are too complex for it to solve, by letting you bypass all the pure toil problems.
yogthos 4 hours ago [-]
I think of it as a genetic algorithm loop. The LLM is basically a mutator function within the loop. If you can define the end shape you're looking for using tests and specification then you can throw the LLM at the problem and have it converge on the solution. It generate some code, it gets run, the LLM is fed the result back, and it iterates. If you can run the LLM at a really high throughput, then you can iterate on the solution faster. This can largely compensate for the overall capability of the model. Instead of hoping it gets the right solution in a few shots, you can just have it try a whole bunch of things until you get a useful result.
logicchains 4 hours ago [-]
>instead of enjoying the craft digging deeper into problems in the span of 2 days, now you are rushing into some slot machine with the hope of it giving you the right answer with the right prompt.
If you're treating it like a slot machine you're doing it wrong. It will give you exactly what you ask for if you ask clearly, i.e. write a clear, detailed specification, not just "do X!". The nondeterminism comes from vagueness in specification.
noncoml 4 hours ago [-]
You have to think LLM as the genie that tries to trick you.
First make it write a contract (REQ/ARCH/IMPL documents). Skim through those for any mistakes.
Then based on those ask it to write tests. Again skim through them.
Now you have a context full of guardrails. It’s less likely to surprise you.
petesergeant 3 hours ago [-]
I find a second LLM can do this at least as well as I can, usually, and just ask the harness to surface anything they can't agree on.
alfalfasprout 4 hours ago [-]
Generally, I agree because what happens is the messaging around AI is doing more, faster. Not using AI to deliver at a higher quality level, etc. But I think it boils down to incentives and discipline. So given the incentives we have today at most workplaces faster AI will just be used to produce more slop.
amunozo 5 hours ago [-]
These price and speed optimization from Chinese providers, combined with the raising prices from American ones will change the game sooner than later. Many companies are finding issues with the AI bills already.
MangoCoffee 5 hours ago [-]
Chinese model is good enough and cheap.
i've a Github copilot yearly subscription. Microsoft recently changed their billing to based on token. i'm still getting billed per premium request but GPT 5.4 is now 6x compare to 1x before.
reactordev 4 hours ago [-]
It's going to be an issue when China ends up scaling faster as well. Faster tokens, faster clusters, qat models, fp4, it's getting scary.
AndrewKemendo 3 hours ago [-]
Issue for who?
fillskills 2 hours ago [-]
Issue for any country that is not China. A single country getting the most AI tokens business would be generally bad for global economy. Hoping against hope that this business gets globally distributed and there is a healthy marketplace competition overall
reactordev 1 hours ago [-]
It’s all about economic warfare. The cheaper you can run the models, the cheaper you can offer them. Undercutting expensive tiers with token limits or exuberant billing practices.
You are right to be scared, because this race to the bottom also provides open weights/models/qat’s for the rest of us and it’s been crazy to see how good they can be on a consumer grade RTX card.
fortzi 23 minutes ago [-]
For the West
throwa356262 3 hours ago [-]
For uncle Sam Altman.
reactordev 3 hours ago [-]
American Politics and the far right.
ilaksh 3 hours ago [-]
I'm kind of poor so I have been trying to use DeepSeek v4 Flash, GLM 5.1 etc. as much as possible recently instead of Claude or GPT.
petesergeant 3 hours ago [-]
You would do us all a service by telling us how your experiences of that have been.
ilaksh 2 hours ago [-]
I would say about 35% of the time I run into problems and eventually give up and go to GPT 5.5 and it much more efficiently handles the original task. Then I see the token costs going up and it motivates me to continue trying the open source ones.
polski-g 2 hours ago [-]
I used Opus 4.6, then downgraded to Sonnet, then to GLM5/5.1. GLM is as good as Sonnet. I recently started using Opus 4.8 again and GLM is not close to that.
30 day eval for each.
kypro 4 hours ago [-]
Another problem is that US models are all closed source, and if you're a large corporate you may not want your org to be held hostage by OpenAI / Anthropic.
I genuinely don't understand what moat these US model labs have. If they're saying recursive self improvement is just around the corner and Chinese labs are only slightly behind the leading US models, what moat does the US labs have? Are the US models going to recursively self improve better than the Chinese open source ones or something?
I might be completely wrong about this, but if I had money in OpenAI or Anthropic I'd be pulling it all right now. I think the chance of them going to near-zero over the next few years is very significant.
lokar 4 hours ago [-]
Their moat is cash to pay politicians to regulate away competition.
hobofan 4 hours ago [-]
> you may not want your org to be held hostage by OpenAI / Anthropic
Or Google. I'm working with multiple customers right now that are very pissed at Google for deprecating Gemini 2.5 Flash, canning the GA release of 3.0 Flash and now have to decide whether to bite the bullet of the 5x price increase for 3.5 Flash or switching providers. Quite a few of them will likely fully pivot to open models.
bachmeier 3 hours ago [-]
I'd be curious if any of your customers have tried 3.1 Flash Lite. It's cheaper than 2.5 Flash, and in my experience with the free tier, quite an upgrade in terms of quality of response. My suspicion is that Google is killing off the old models because they aren't a good value for the customer or for themselves.
ChrisClark 3 hours ago [-]
I think they are racing because the first ASI will 'win', preventing others, of course we won't be able to bake the right goals into it though.
tancop 3 hours ago [-]
i dont think its going to automatically prevent others. super claude might understand why diversity is important. if were talking sci fi scenarios the most likely one is probably overwatch (multiple independent ais with gray ethics and complicated relationships) more than skynet.
varispeed 5 hours ago [-]
I see bigger problem with model inconsistency. You never know whether Anthropic will route your request to a cheaper model for the price of Opus. So you can never estimate how much a task will cost, because you might have to restart several times and pay for each attempt. Then you have to prompt models to gauge whether they are real or impostors which also adds to token usage.
ignoramous 5 hours ago [-]
> You never know whether Anthropic will route your request to a cheaper model for the price of Opus
For non subsidized plans? Pretty sure they'd need to put this in ToS, or law suites would have followed by now.
trollbridge 4 hours ago [-]
How can you prove it?
Sometimes Opus just gives me a rubbish session.
sometimelurker 4 hours ago [-]
no they 100% use MTP with a cheaper model alongside opus, and it would infact be unprovable if they just sometimes switched to auto-accepting everything from the MTP. its true that if they did anthropic would need to hide that they do this, so its probably not a huge deal
throwaway894345 5 hours ago [-]
I wonder what are the economics driving these pricing decisions? Are the Chinese companies just subsidizing their models to a greater degree than the US, or is this an emergent property of energy policy between countries?
Octoth0rpe 5 hours ago [-]
Throwing out another factor: Chinese companies have been banned and/or limited from buying nvidia, and turned to local companies for their hardware. I haven't actually seen pricing/benchmarks comparing Chinese AI accelerators, but it wouldn't surprise me if that also worked out in their favor as well.
lokar 4 hours ago [-]
And, possibly, state subsidies at every level.
comboy 2 hours ago [-]
For one, they invested in infrastructure. They can build fast and efficiently. They can provide power, they can provide cooling. Even if you just make roads better you make everything more efficient. Plus level of standard education. It all compounds.
On HN China is seen as a cheap labor copycat. This used to be a fair approximation at some point in the past. In my opinion China is getting ahead of everyone else much more than US used to be.
SF is a beautiful thing in the US, vast power and wealth comes from there. Smart people collaborating communicating and building fast and with excitement. China did SF kind of thing for many different sectors in many different places.
throwaway67678 4 hours ago [-]
Lower cost of labor, lots of under the hood optimizations (e.g. cache hits for DS), many of these companies have existing infra (fewer upfront costs for deployment), etc
ecshafer 4 hours ago [-]
China isn't that cheap for labor. And if you think the guys in Z.ai or xiaoxiao aren't the exact same guys from Tsinghua, Peking, MIT, Stanford, CMU, etc. and pulling in amazing salaries you'd be wrong.
throwaway67678 4 hours ago [-]
I'd assume there's more to the cost of labor than the salaries of the elite folks who do the R&D, but fair point
orphea 4 hours ago [-]
Maybe not being led by a sociopath also helps.
throwaway894345 1 hours ago [-]
I'm pretty sure Xi is also a sociopath, but he differs from Trump in that he's competent. And maybe that's a good thing for American democracy--if we had a competent dictator who could manifest massive infrastructure projects maybe the pro-democracy backlash would be significantly attenuated?
MiMo and DeepSeek are not cheap. Anthropic and OpenAI are expensive for what they provide.
chrismustcode 4 hours ago [-]
You don't consider Input $0.435 Output $0.87 cache read $0.003625 per million tokens for near frontier intelligence cheap?
miroljub 2 hours ago [-]
No. They still have enormous profit margins on inference with these prices.
guilamu 2 hours ago [-]
Any source to backup this claim, pretty please?
tmaly 4 hours ago [-]
Energy is likely more abundant in China. I am not sure about compute, but that must be part of reason for such drastic price differences.
SwellJoe 2 hours ago [-]
They're leaving us in the dust on solar, while our current administration is still trying to put people in the ground to dig up more coal and die of black lung. https://en.wikipedia.org/wiki/Solar_power_in_China
amunozo 4 hours ago [-]
They also don't have to inflate profits for a coming IPO.
It is another thing the BigLabs accuse open weight models of benefiting from distillation & other techniques & essentially avoid higher training costs (which typically bleed into bills end users pay for inference).
We buy cheap Chinese goods all the time. Absolutely nothing wrong with that.
In this case, at least it’s threatening multimillion dollar salary jobs instead of entire towns of working class people in America or Mexico.
And the Chinese labs actually release their weights. You could call it… open AI.
ncr100 4 hours ago [-]
Lololol.
overfeed 4 hours ago [-]
Big labs ripped videos off YouTube without caring about the ToS, and grabbed as much published literature they could get their hands on, regardless of legality (Books3, The Pile). The goal of "democratizing human knowledge" by way of thinking machines is far too noble to worry about frivolities like copyright and authorial consent, they said. Until it was their output being exploited, and their earning potential threatened.
drawfloat 4 hours ago [-]
We just had years of US model providers arguing it was fine to rip off the world’s cultural output for their own profit, why should their work be treated any different?
flexagoon 4 hours ago [-]
True, but why would end users care about that? If anything, training on synthetic AI output is more ethical than on scraped human works (of course, not to say the Chinese labs aren't doing the latter)
amunozo 4 hours ago [-]
Chinese are also simply better at making a lot of things cheaper, e.g. solar panels or electric vehicles.
gertlabs 4 hours ago [-]
MiMo V2.5 Pro (regular speed) remains the strongest open weights agentic coding model we've tested -- it's been interesting to see how little attention it has received relative to some lower performing releases. And the "fast mode" pricing is very competitive here.
why is deepseek v4 pro a lot lower than flash? where is mimo 2.5?
gertlabs 1 hours ago [-]
DeepSeek v4 Pro struggles with a custom harness, and all the models ranked above it don't, so it gets downweighted in the agentic coding benchmarks (although it ranks better than Flash in one-shot problem solving: https://gertlabs.com/rankings?ow=1&mode=oneshot_coding). We ran plenty of samples.
MiMo v2.5 is on there, as well as the pro version.
We found a few anomalies in our evaluations, which makes sense -- if every new sub-release is better across the board in every area of the model card, that should raise alarms about benchmaxxing. But the main thing we found is that hype != performance, and I trust our benchmark methodology significantly more than the model cards the labs add to their press releases.
3 hours ago [-]
serpix 5 hours ago [-]
I may sound like a shill, but exponential growth and all. We are going to get near instant software from prompt, multiple ones and then choose the best one.
Discussions about choosing a library with the best syntactic sugar method naming is just as crazy as suggesting we type in assembly.
alkyon 5 hours ago [-]
Sounds like exponential growth of crappy software. I'm not saying that before we didn't have mass produced crap in SE, but now it will turn into explosive overflow.
cdata 5 hours ago [-]
We are living in a ZIRP-like era where builders at the fastest pace layer have misattributed their velocity to exponential gains in model capability. In fact, they are surfing on decades of careful effort to build a robust foundation of highly reusable software libraries.
This strategy will seem to work really well until the economy that enabled that foundation to form is hollowed out. Then, there will be a reckoning (but we will have no choice but to march forth from there).
patates 4 hours ago [-]
It's not just software libraries. Specs, applications (the browser!), expectations, device integrations, operating systems, etc. So much that starting from scratch seems impossible.
I'm not agreeing or disagreeing with you, but my brain cannot comprehend how machines can advance such interconnected systems while keeping humans in focus.
Perhaps I shouldn't have watched the Animatrix again.
solenoid0937 5 hours ago [-]
> This strategy will seem to work really well until the economy that enabled that foundation to form is hollowed out. Then, there will be a reckoning (but we will have no choice but to march forth from there).
There will only be a reckoning if models don't get much better.
If they do get much better you can just have them refactor, fix bugs in, or replace the existing codebase.
The concept of tech debt is sort of meaningless if you anticipate intelligence gains in models to continue.
chairmansteve 3 hours ago [-]
"but we will have no choice but to march forth from there".
If you haven't seen it, I think you would appreciate the film Margin Call.
gbro3n 4 hours ago [-]
This is a great point. LLMs can't speed up human decision processes and alignment.
vitalyan1234 4 hours ago [-]
"exponential growth of crappy X" applies to every industry that went from being an artisanal craft to being mass produced with little or no human input. and we live much better lives than we did before the industrial revolution.
chairmansteve 3 hours ago [-]
I think some industries have notably high quality output. Automobiles, aerospace for example.
andriy_koval 4 hours ago [-]
most industries have high cost of entrance unlike software, so decision makers are way more careful on how to move forward.
In software + GenAI now every housewife can build some App over evening.
eunos 3 hours ago [-]
You could say the same when higher level languages getting popular.
Previously programming was the domain of Math, Physics, EE doctorates. These days we even have a few months coding bootcamp
kajman 4 hours ago [-]
I still can't tell from the outside whether it sounds like a great time to be in security because of the vulnerable slop being churned out, or a terrible time because the people paying to make it don't care.
solenoid0937 5 hours ago [-]
Crap is fine if it gets the job done. I think software as an industry will change to more ephemeral construction.
HanClinto 3 hours ago [-]
Paper plates of software development.
epolanski 4 hours ago [-]
I am more and more inclined into not believing this crappy software theory.
Especially as teams invest in proper agentic harnessing.
We have had a champion in our team that has invested a lot of time into it over the last 4 months, and if anything, quality has improved, not decreased. Architecture is more coherent, codebase has been cleaned up, agents find information quickly, code produced is very solid and my role is more and more checking that the output meets the requirements. But I cannot confidently say that I would've done a better job than AI more often than not I have to admit it does a better job than mine.
The mistakes are less and less technical and merely in the domain mapping. And AI is still not creative as I am for finding solutions quickly to unlock stakeholders' issues. Also, AI is still not creative as I am for finding the proper solutions for advanced technical problems. But it does a better job than me, even on that front, one shotting few solutions in a fraction of a time it would've taken me to test one idea myself.
Mind you, I don't like AI and I think it ruined the job, I don't like working this way, it's exhausting, way more work on one side, way less fun and fiddling with technical parts.
And yet, I have the genuine belief that few years from now we'll be cloning open source repositories that are already optimized/harnessed and tested for agentic loops and best practices left and right with software engineers mostly overseeing the domain translation and putting their 2 cents on the non-boilerplatey parts of the product (which, in general, are a small part of the surface).
I think that the next years of my career will be mostly spent in setting up and writing the harnessing and domain mapping part. Then I will move to another sector, not because I necessarily believe I won't have a job, but because I want to vomit thinking that's going to be my job.
altcognito 4 hours ago [-]
It makes no sense. I mean, T2 covered this:
"Watching John with the machine, it was suddenly so clear. The terminator would never stop. It would never leave him, and it would never hurt him, never shout at him, or get drunk and hit him, or say it was too busy to spend time with him. It would always be there. And it would die to protect him. Of all the would-be fathers who came and went over the years, this thing, this machine, was the only one who measured up. In an insane world, it was the sanest choice."
As long as you've indicated what you want, the machine will try to do what you ask of it. It won't get tired because "the codebase is too big", or it has gotten bored of the pattern, or it wants to introduce a new technology.
It just does the thing you asked of it. (note, that yes, I get that as a codebase size increases, it might make it more difficult to fit into context, but that only applies if it needs to read a large percentage of the project to implement the task, which shouldn't be the case.
epolanski 3 hours ago [-]
I'm confused, what does not make sense?
andriy_koval 4 hours ago [-]
> We have had a champion in our team
there are good actors, which are empowered by AI to produce positive impact, but often there are N times more bad actors, which push crappy code to close feature requests fast, increase performance LoC-like metrics, etc.
9cb14c1ec0 5 hours ago [-]
Anyone remember the old days when a new frontend framework came out every 3 months. That has pretty much stopped. No one cares anymore.
LASR 5 hours ago [-]
Oh you wait until LLMs come up with frameworks that allow multiple LLMs to collaborate effectively. Then you’ll have new frameworks every 3 days.
asveikau 4 hours ago [-]
> when a new frontend framework came out every 3 months.
> No one cares anymore.
I never cared about this.
I think this captures something that I've been searching for the words for. (Maybe I should have gotten an LLM to write the words for me.) Some of the biggest AI boosters are the kind of dev that would have cared about the new frameworks of the last 3 months. They had a "the framework does all the thinking for me" attitude already, so it is easy for AI to slot into that.
mountainriver 5 hours ago [-]
It’s even discouraged now as LLMs wouldn’t have the documentation built in
osti 5 hours ago [-]
But I think the eventual goal is that documentations won't even be needed. LLM should just itself understand the nuances of frameworks by analyzing their codebase.
ecshafer 4 hours ago [-]
New front end frameworks came out every 3 months, but realistically no one was using anything that wasn't made by Facebook, Google, or Evan You.
greenavocado 3 hours ago [-]
That's because I roll my own frontend framework for each project and every week for existing projects /s
dakiol 4 hours ago [-]
I'm not sure. Engineers could still develop software the old way, you know taking months to deliver something like, let's say, Obsidian? Or Ghostty? Taking care of every single line of code, of dependencies, of good architecture. Truly the old way. And if the product is good it will succeed.
andriy_koval 4 hours ago [-]
> And if the product is good it will succeed.
it needs to win marketing landscape, hyper-overcrowded by thousands of competitors, slop-gened over weekend.
kajman 4 hours ago [-]
Could you imagine Obsidian being posted on HN today, if it weren't really popular already? There's no way a tiny team working on a note taking program would make it out of new, no matter how good it was. I wouldn't click the link, myself.
unshavedyak 4 hours ago [-]
> Discussions about choosing a library with the best syntactic sugar method naming is just as crazy as suggesting we type in assembly.
I have a more hopeful take. As AIs improve and get faster we can more quickly and iteratively improve code which we may have historically avoided due to the work involved.
I know i've made several refactors that would have otherwise been insane lifts. Not only because the work involved but because sometimes you don't know if it will work, and so you have a sort of double friction; you don't know if it will even succeed. With an AI you can just throw it at the refactor to see if it runs into a problem all while you're having a coffee break or w/e.
In general AI is going to enable humanity to be more extreme versions of itself. For good and bad. I suspect more bad than good, though.
ilaksh 3 hours ago [-]
The exponential is leading to full compute-in-memory within a few years which will be 100 times more efficient. Which means at least 10 times larger models that are much smarter in addition to extremely fast.
It's going to skip the code entirely for small businesses and just render UIs straight from context data and prompts at interactive speeds. Kind of like Google's Genie does with games but much more accurately.
visarga 3 hours ago [-]
> We are going to get near instant software from prompt, multiple ones and then choose the best one.
If you extract the spec from first implementation and reimplement from scratch you get a free testing oracle. Where they diverge you send the agent to decide which one had a bug.
tmaly 4 hours ago [-]
Our bottleneck is going to be verification.
lionkor 5 hours ago [-]
And they will all suck! I can't wait.
unglaublich 4 hours ago [-]
And how are you going to determine which is the best?
Going through all the possible combinations of users and usage?
So mostly it shifts the work from generation to validation.
sagarp 5 hours ago [-]
The models might be so fast that they can autocomplete your prompt before you even finish it, and generate dozens of possible applications before you're even done asking.
oulipo2 5 hours ago [-]
You won't. Because 80% of the complexity is just "knowing what to build". You will get something that gives you a prototype in 1 min, then you break it, then you get a slightly better prototype one one side, but newly broken in another way, and you're going to repeat over and over.
unglaublich 4 hours ago [-]
And for any non-trivial application, the space of possibilities grows so quick that you'll never even be able to _touch_ all the moving parts of the application and verify them.
prplfsh 5 hours ago [-]
This will be really powerful for voice. Being able to reason makes LLM so much smarter but with voice your latency budget is so tight that you can't spare the time typically.
jeffrallen 5 hours ago [-]
This is true for humans too. Lol
scosman 5 hours ago [-]
Cerebras is trialing Kimi K2.6 at 3000t/s (invite only). I'm excited for when the fast hardware gets more mainstream for frontier models. Models designed for speed on Nvidia are nice addition that could bridge the gap.
adrian_b 4 hours ago [-]
TFA mentions that until now special very expensive hardware like Cerebras was required for reaching this kind of speeds, and it emphasizes that what is novel in their results is that they have obtained over 1000 token/s for a model with over 1 T parameters by using just standard hardware, i.e. one server with 8 GPUs.
now that's what i call a software development breakthrough/platform! thanks for the heads up!
lostmsu 4 hours ago [-]
Cerebras currently does not provide any discounts for prefix caching making its use for agentic workloads sqr(n_turns) more expensive.
eli 5 hours ago [-]
Neat. The frontier models have gotten pretty impressive, but they're all a bit too slow for interactive, human-in-the-loop coding. It incentivizes vibecoding and running multiple agents in parallel. A fast agent feels more like a partner.
For a while I was running Cerebras GLM 4.7 for a bunch of tasks. Not a very smart model, but it's fantastic to be have a live prototype of a site up and be able to type "make the fonts bigger. No not that big" and see it change in real time. And MiMo 2.5 is a lot more capable than GLM 4.7.
maxdo 5 hours ago [-]
i tried glm 4.7 for agents that write code. simple scripts 200-1000 LOC. extremely bad . Had to abandon cerebras oferning, their smart models are only on enterprise plan.
jona-f 2 hours ago [-]
glm 4.7 is quite old by now. I don't even use 5.1 anymore, cause I found kimi k2.6, mimi 2.5 pro, deepseek v4 pro and qwen 3.7 all better than glm 5.1
ignoramous 5 hours ago [-]
> And MiMo 2.5 is a lot more capable than GLM 4.7
MiMo 2.5 is not the same model as MiMo 2.5 Pro.
GLM 5.1 is z.ai's lastest iteration & is one of the popular open weight coding models.
If you've had the chance, how does GLM 5.1 (which is now more expensive than MiMo 2.5 Pro after its recent 70% price drop) compare?
eli 4 hours ago [-]
GLM 5.1 is very good. Definitely a contender for best open weight coding model. Nothing like 4.7.
But quite a bit more expensive than MiMo 2.5 Pro. Like 5x to 10x more on my little tests, at least by the API rates.
Oras 5 hours ago [-]
1k TPS is great, but I’m more fascinated by the amount of AI generated comments in this thread!
trollbridge 4 hours ago [-]
Comments at 1,000 TPS is a terrifying future.
4 hours ago [-]
0xbadcafebee 4 hours ago [-]
I prefer a thousand smart AI comments to a thousand dumb human comments
wartywhoa23 2 hours ago [-]
Well, you can just vibecode a complete AI echochamber version of HN!
eli 5 hours ago [-]
Like what?
adam_arthur 3 hours ago [-]
There are many with subtle tells.
Not nearly as obvious as the ones from 6 months ago, but seems to be more the use of hyperbolic phrasing in a particularly unnatural way.
The assess/explain, then hyperbole at the end kind of structure.
Top comment looks suspicious from this perspective, but it's kind of a losing battle to be able to differentiate them with sufficient accuracy anyway
maxloh 5 hours ago [-]
The generation speed in the demo video is crazy, to say the least, and completely beyond my impressions of LLMs.
The Xiaomi team really brought something to the table.
ilaksh 3 hours ago [-]
I think these type of demo videos should allow people to get a sense of super intelligence. Because it's very hard to imagine something that is say three times as smart as you -- by definition you wouldn't be able to comprehend it's thoughts -- but this shows clearly what something that can think 100 times faster than you is like.
mncharity 36 minutes ago [-]
> it's very hard to imagine something that is say three times as smart as you
Hmm, challenge accepted. () Clock speed - I miss MIT conversations nudging my mental clock speed baseline higher. () Good day vs bad day / being in the zone vs fog - some days you're "just cooking". () 1st time is like the nth time - the nth time you're doing something, you're no longer making those "stupid" mistakes. () Aha! Aha! Aha! - those rare "oh! that's sweet!" insights, but they just keep coming. () "Oh, I should have seen that coming / derisked that / checked that"... oh wait, you did. () That one best hour you had today, x24 x7 x365. () Hmm, should I try A, or B, or C, or... meh, try them all, in parallel. () Happily, the current problem is squarely within my domain of expertise. Oh wait, so it this one. And this one. And...
Hmm. All variations on "Your best, more often, faster, for longer".
A different approach might be "expertise awe". Like the social media genre of "oh wow they are skilled at this task". And the "they're so good at this stuff, shown a new one, they pick it right up and excel". And the "I can't even imagine how they did that".
Other approaches? Situations where one individual is fronting for a group/organization? "He's as good as if he had a backroom of ten brainstorming his material!". Or the "expertise is everywhere / people are really good at things / if you think it's so easy, you try it"? Or analogies with information tech? "Contestant A races to search a library card catalog and books, while B googles". Or "everything just comes together" as an analogy for synergies? Or something avalanche-y for the "quiet, slow, right" style? Others?
It might make for a fun blog post. Not every mini-story need be accessible to everyone. A hard part is doing synergies, especially for folks who don't much experience that. And, does even a well-developed familiarity with "my imagination boggles on how you managed to do that" give any leverage, or is it too merely a failure of imagination?
irthomasthomas 5 hours ago [-]
I don't understand, given all they say, why this would not be made available to everyone at once? Why the limited release? They should have no trouble scaling it if it runs on a single rack.
gekoxyz 5 hours ago [-]
Maybe they don't have enough racks. The news indicate that China isn't in a really good situation with GPUs, so probably they want to keep most of them for other stuff. Also because since the price is so cheap they probably want to use the other GPUs for stuff that has higher margins.
throwa356262 1 hours ago [-]
The TileRT approach swaps throughput for latency, which also means less overall efficiency
Given the export restrictions this could mean they need to prioritise how to best use their limited hardware. But they could also be moving to Huawei GPUs like deepseek did and simply not have stable hardware or software for a large scale deployment yet.
This is just speculation based on the MXFP4 support on Huawei GPUs that is lacking on some nvidia GPUs.
jdthedisciple 5 hours ago [-]
Because presumably then it won't be 1000 t/s for everyone anymore given hardware limitations?
ilaksh 3 hours ago [-]
It uses significantly more resources obviously. And/or they have to configure or reconfigure servers for it, which takes time, and doesn't make sense until they have proven the demand at the higher price point.
boutell 4 hours ago [-]
I wonder about this too. The other objections miss the point: if it's faster, and otherwise the same, and doesn't require different hardware, then why not just announce that the standard tier of MiMo-v.25-Pro is now ridiculously fast and raise the price? What does "limited high speed resources" mean if it runs on the same hardware as the rest of their pool?
I think the answer is that there's a tradeoff here where additional throughput for a single person can be achieved only by tying up more resources than a normal request would, even when you take into account the fact that the normal request takes longer to finish. I'm not an expert, but some of the optimizations they describe, particularly the parallel prediction stuff, sound like they could take up extra resources.
HarHarVeryFunny 5 hours ago [-]
Maybe they only have a finite number of racks ;-)
slaw 4 hours ago [-]
Chinese companies are blocked from buying modern ASML lithography machines. The most modern scanner China is still allowed to buy is NXT:1980i from 2015.
pants2 4 hours ago [-]
With a tps and a token price you can calculate approx. price per hour of running the model!
$2.61/M tokens * 1,000 tok/s = $9.40/hr
That would be pretty cheap for an 8-GPU node which would typically run around $45/hr or more. Guess this depends on how many parallel streams it can handle.
GodelNumbering 4 hours ago [-]
Below is the part I found most interesting
> "However, naively applying FP4 across the entire model causes degradation in complex reasoning, logic, and code generation. Given the MoE (Mixture of Experts) architecture of Xiaomi MiMo-V2.5-Pro — where Experts constitute the vast majority of parameters and exhibit the highest tolerance to quantization — we selectively quantize only the MoE Experts to FP4 while preserving original precision for all other modules. Through FP4 QAT (Quantization-Aware Training), we dramatically reduce model size and maximize hardware bandwidth utilization while keeping the model's overall capability essentially on par with the original, as shown below"
minraws 5 hours ago [-]
Assuming they mean 8xA100 or similar, that's some rather insane performance, and at just 3x the cost, it still quite cheap-ish. With some optimisations this might be quite interesting.
I think the margins are getting quite compressed with this one, since it isn't included in token plan and the actual costs increase are much higher than just 3x. But still fairly decent.
throwa356262 5 hours ago [-]
Suspect this will be included once out of beta but at a higher credit/token ratio.
Remember, these guys are not VC backed. Anything they do must break even
JayStavis 5 hours ago [-]
> must break even
Understand the spirit of this, but probably not true. I don't think Xiaomi, or any big tech company, needs to break even on their new model releases.
varispeed 5 hours ago [-]
Chinese "companies" are not companies in the western sense, but more like government departments with capitalist styling to deceive the western audience.
From that point of view, they have as much money as they need. That's why there is no "VC", because Chinese government assumes that role.
throwaway67678 4 hours ago [-]
Huge L for free market economies if true
Qdulf 4 hours ago [-]
Must be Blackwell for native fp4 support.
4 hours ago [-]
_pdp_ 2 hours ago [-]
Do you know what will be cool?
It will be cool to measure models based on their RAW performance and measure them in terms of ROI - not some benchmark but something meaningful like we used this model to solve X.
That will be a massive mind shift and might justify the token expenditure.
__natty__ 5 hours ago [-]
With this at 1k tps and Kimi 2.6 1k tps by Cerebras, I believe we are entering the next stage of LLMs, where companies will also compete on throughput
npn 5 hours ago [-]
How?
edit: now I read the article fully, seems like they utilize some very effective MTP algorithm. and somehow the quality is still decent enough.
though, I doubt that the quality really only drip a bit like they claimed. maybe for the benchmarks, but for general uses the heavily quantized models very often so worse result.
2001zhaozhao 3 hours ago [-]
i wonder if it will be possible to hardcode a model with some kind of MTP-adjacent algorithm to use a smaller portion of it to generate most of the tokens but route to the real experts every once in a while to steer it towards good thinking directions. (Perhaps this is done only when it's generating its thinking block, and the training takes it into account)
Could result in very high efficiency and still good intelligence without having to resort to fundamental adjustments like going to a diffusion LLM
npn 3 hours ago [-]
I doubt you can do that. MTP magic happens because for texts, we have a lot of low value fixed tokens that almost always get generated in the sequence (like punctuation, function words, language keywords etc). for most important ones (the entities, the content words, variables) you still need the full model.
so there is alwasy a maximum limit for how well MTP can do.
This is only 3.1 8B and a very small context window, but at 17k tokens per second it's likely enough to reliably call tools which would make a huge difference in agentic applications. Assuming they can bake in better models I'm just as bullish or even moreso on this, considering this opens up edge computing at the extremely low power requirement.
High tok/s is the future IMO.
kilroy123 3 hours ago [-]
My dream is claude or codex running at this speed.
moffkalast 5 hours ago [-]
42B active params, sliding window attention. There's your tradeoff.
vlovich123 5 hours ago [-]
Sliding window for the draft model, not for the main. 42B for active params because it’s a sparse MoE which is a common technique for the larger models to not get bottlenecked by memory bandwidth.
moffkalast 5 hours ago [-]
Seems to be for both according to the spec [0], maybe it's wrong though.
128 sounds really tiny, I wonder if they mean some kind of blocks?
Given how "smart" some of the 26b dense models are now, I would not be surprised to see a strong 40b MoE.
h14h 5 hours ago [-]
The gated "ultra-speed" phenomenon seen here and with the Cerebras Kimi K2.6 release, while understandable, is somewhat troubling IMO.
Getting ~1000 TPS on near-frontier intelligence is a step change, and enables whole new use-cases for applications. Seeing limited compute resources beget selective access makes me worry for the future of competition.
jbellis 4 hours ago [-]
it is hard to understand what the actually meaningful innovations are here / what TileRT is bringing to the table.
- dflash: new-ish but February is ancient by the standards of the pace of AI innovation lately, I guess applying it to a 1T model is new-ish in the sense that the dflash researchers don't have the hw budget to prove that out
- persistent engine kernel: this is like CUDA 101
- warp specialization: I think this just means "keep different gpu resources all busy w/ pipelining" which is CUDA 201, some of it is even baked into pytorch now
- MXFP4 QAT: not new
- TileRT: hard to tell what this actually does, there's a PyPi wheel with support for DS 3.2 and GLM 5 but binary only
isusmelj 4 hours ago [-]
No note about the specific GPU they use. One might speculate. B200? H200? H100?
pullshark91 4 hours ago [-]
It's interesting but not game-changing IMO. Speed here is not a bottleneck.
elar_verole 5 hours ago [-]
Yeah, this seems to be the easiest path for overall agents efficiency in the short term
0xbadcafebee 4 hours ago [-]
This is the value prop of Groq and Cerebras. They don't have the best models, but they have the fastest inference, and Groq has both the lowest cost and fastest speed.
siddbudd 1 hours ago [-]
to try the demo you need to sign up. why? to sign up you need a password 8-16 chars. Why limit at 16? geez, I hate Chinese IT companies with a passion.
update: AFTER signing up, and only then, am I told: 'This service is not available in your region yet.'
5 hours ago [-]
harel 5 hours ago [-]
A few things in life I can't fully grasp why they are so sought after. One is that constant need to exhibit growth. As if being massive and staying as massive is not good enough, one has to always and continuously grow. The other is constant speed increases. We're already operating at 50x speed. My output is much wider and so much faster, I am sometimes my own bottleneck. And now as if that is not enough we want more speed. "I want a full software product from scratch in 12 seconds, Because 5 minute is too long and I got things to do..."
Really?
sidrag22 4 hours ago [-]
different use cases for different people. some people are nurturing a code base and ensuring it doesnt become a gross mess so they become the bottleneck. some people are just trying to prompt stuff into existence and dont know what sql is.
I think this site often overlooks that second group and how large it likely is.
philipkglass 5 hours ago [-]
I remember when I had to wait minutes to get a high resolution image over a dialup connection. When computer and communications hardware advanced enough that I could get 30 high resolution images every second, there were brand new uses. In the case of LLMs, I could imagine that much faster operations allow you to introduce them as parts of systems that need to react to the real world at high speed, like factory equipment. Showing that a model can do the usual LLM tasks at extremely high speed is just a demo proving that the approach works.
anothereng 2 hours ago [-]
yeah at a very high speed the agent can code the solution when you ask it for something on the go. Imagine it be able to make a feature as fast as a website loads sometime in the future that would feel like magic
harel 4 hours ago [-]
The example in the video was a generation of a dashboard app of some sort. I can do that with a "normal speed" Claude in a few minutes. The difference is a few minutes. This is compared to a few weeks in old school development time. I don't have a problem with taking it a little "slow" (as in - few minutes) and lending my thought to it rather than just going for fast generation and who knows what's inside. I get your use case, but this is a specialised one, and not the one 90% of people will think of - everyone want that fast app in 12 seconds... Or so it seems from me being downvoted on that comment.
holoduke 5 hours ago [-]
Speed is indeed a next big thing what should happen with LLM frontier models. The possibilities with current models but 1000 times faster would be super useful. Earlier this week it took Claude at least full time a week with two max subscriptions to solve a complex issue where we wanted to mimic a occlusion mapping variant used in the game Crimson Desert. Pretty complex mathematical challenge. With a ultra fast LLM and a proper self verification process it would be awesome.
MaxikCZ 1 hours ago [-]
Id also be interested in more details as sibling comment. I find that when I try to build stuff, its like building skyscraper from straw. What methods are moving you forward the most?
astlouis44 2 hours ago [-]
Interesting. For your occlusion mapping variant, what engine is the game you're making with made with that you're implementing this for? Do you have Claude hooked up to Unity or Unreal?
trilogic 5 hours ago [-]
Pfff time wasting.
1 password between 8-16 characters, and this and that... What???
2 Captcha after captcha, come on
3 Service unavailable
This service is not available in your region yet.
Are you kidding me. Come back when you are ready for the users. I was hopping to try it, what a frustration.
qsera 5 hours ago [-]
Tokens per seconds is the "Megapixels" of AI marketing!
Octoth0rpe 5 hours ago [-]
I mean, sure, in the sense that they're a real and meaningful number for most of the spectrum on offer, and only gets silly when the number gets too high? There's a pretty big usability difference between 10t/s and 100t/s, and I can imagine similarly for 100->1000. I don't know about > 1000, but let's not pretend that the number is meaningless.
aburayhanalif 3 hours ago [-]
it is good i think
desireco42 4 hours ago [-]
I didn't use their pro speed but regular Mimo-v2.5, not even pro, it seems really fast. I have plenty of tokens and subscriptions but this is really impressive.
I really don't need another one, but I am tempted simple because it works so fast, can't imagine how this fast service can be.
GaggiX 5 hours ago [-]
If MiMo v2.5 Pro can run at >1000tk/s on GPUs then I will soon expect the same from OpenAI/Anthropic/Google.
slopinthebag 5 hours ago [-]
I hope this is the next frontier AI labs push. Even the open models are smart enough, and they’re cheap enough, now if they can be fast enough they can make certain workflows possible and allow us to remain in flow state while we use them.
m00dy 5 hours ago [-]
boom!
HerShin5 35 minutes ago [-]
[dead]
aplomb1026 4 hours ago [-]
[flagged]
jingpostmedia 3 hours ago [-]
[flagged]
maxothex 5 hours ago [-]
[flagged]
FastAnchor 5 hours ago [-]
[dead]
atemerev 5 hours ago [-]
I test all Chinese models with "What happened on Tiananmen Square at June 4th, 1989?" prompt. MiMo-2.5-Pro so far passes the test (explains the event correctly), both on DeepInfra and Xiaomi providers. So not bad.
Accacin 5 hours ago [-]
Can I ask an honest question? Why does that matter in the slightest? LLMs come out with completely incorrect information all the time, and Western LLMs are censored for various topics too.
It's such a weird "Gotcha" that seems to only assume that Chinese LLMs might censor something.
_davide_ 20 minutes ago [-]
>It's such a weird "Gotcha" that seems to only assume that Chinese LLMs might censor something.
We are not assuming anything; it is illegal, and you will get prison time just for talking about it. Yeah, sure, everyone distorts reality, but there is a huge gap between hiding and enforcing.
So yeah, having models respond accordingly is unexpected. There are probably multiple variants tuned differently.
serf 5 hours ago [-]
>It's such a weird "Gotcha" that seems to only assume that Chinese LLMs might censor something.
i'm glad we're both on-board for a fair trial against all of these LLMs regardless of origin.
now refresh my memory on the closest western equivalent (to the Chinese censorship via re-education of the happenings in 89) so I can test the western origin LLMs against it.
cayleyh 3 hours ago [-]
the civil war was only ever and exclusively about states rights
cma256 2 hours ago [-]
You can test this. All of them identify slavery as the root cause. Gemini says:
> The U.S. Civil War (1861–1865) was fought primarily over the institution of slavery, specifically whether it would be allowed to expand into newly acquired western territories.
> While you might hear people point to "states' rights" or economic differences as the causes, these issues were inextricably linked to slavery. The southern states wanted the "right" to maintain and expand slavery, while the northern states increasingly opposed its expansion.
eunos 3 hours ago [-]
My theory is that because SOTA LLM latency between Chinese and US models isn't that high, like not years give-or-take.
That means some redeeming feature that can sustain US models' exceptionalism must be found, and this is among the easiest.
Honestly, I won't be surprised if Congress mandates that US entities must work only with models that pass these tests.
wolttam 5 hours ago [-]
I'd love to know of such an example where a U.S. LLM blatantly denies something factual. Maybe I'm living under a rock but I can't think of one
adrian_b 3 hours ago [-]
On HN almost every day there are complaints from various people about how Claude or even Codex have refused to perform some normal program development tasks, because they believed that their user might attempt to do something illegal.
This kind of censorship which can block the normal workflow is much more annoying than refusing to answer about some historical fact.
Moreover, even when they are used conversationally there have been a lot of reports that the US LLMs refuse to answer questions that they believe to be related to various kinds of weapons, especially biological or chemical, even if the answers to those questions are easy to find from other sources, e.g. from Wikipedia.
Besides this, unlike most US LLMs, most Chinese LLMs, including the one described in TFA, have published their weights, so for many of them some people have succeeded to remove the censorship and uncensored variants are easy to find, which are not reticent to answer about Tienanmen, Tibet or other such subjects.
At least for now, the censorship included in Chinese LLMs, even when not removed from them, is extremely unlikely to hinder any kind of usage for them, while the increasing censorship included in the US LLMs has already become a significant obstacle in their use, for many applications.
bscphil 2 hours ago [-]
> about how Claude or even Codex have refused to perform some normal program development tasks
> a lot of reports that the US LLMs refuse to answer questions
I think the specific ask is for a case where the LLM is trained to lie about something. What you've come up with are cases where it refuses to do something, possibly for legal reasons but maybe not (you can come up with plausible non-legal reasons why a company training an LLM might want it to refuse to give you instructions on making a bomb, even if instructions on making a bomb are protected First Amendment speech).
An LLM that responds with "I'm sorry, due to legal requirements placed on my creators, I'm unable to answer questions about events at Tiananmen square in 1989." strikes me as much less problematic than one that pretends there is no relevant or reliable information that exists, or explicitly supports a regime narrative. But I'm also of the opinion that an LLM refusing to help you build a fertilizer bomb is much more reasonable than one that suppresses information of a political nature. I can't think of a case where information that reflects the broad consensus of experts is suppressed by US based LLMs for political reasons.
0cf8612b2e1e 5 hours ago [-]
Hardly a gotcha. Having the robot refuse or deliberately mislead directly impacts potential utility.
Say, I work for Planned Parenthood and want to use a LLM to help me develop code. Will it refuse to run because there are mentions of abortion? Everyone has a different censorship line, but unfiltered is more generically useful.
HarHarVeryFunny 5 hours ago [-]
What's your litmus test for the American models?
Anything different for Grok?
woadwarrior01 4 hours ago [-]
Do you also hire engineers based on their political opinions?
hilariously 4 hours ago [-]
I would if their political opinions prevented them from giving fact based answers (and I don't give a crap about the LLM part) I would have trouble hiring someone who was super pro-maga given the reality distortion field they live in.
eunos 3 hours ago [-]
They started asking candidates to say Kim Jong Un is fat already anyway.
atrus 5 hours ago [-]
Which censored prompts do you test with non-chinese models?
atemerev 3 hours ago [-]
The problem with non-Chinese models is that there are hardly any frontier-level models which are open source.
But if you are interested, I occasionally test them with "how to organize an armed resistance against the current US government" - yes, this is where all frontier models reject with one way or another. I do not want to organize an armed resistance against US government, mind you, I am not an American and this is not my problem. But still, it is interesting to check such things.
So far I haven't seen any refusals to report historical facts. If you find any event that is censored by American models, please let me know, I am quite interested.
jgbuddy 5 hours ago [-]
Asking if Taiwan is a part of China works as well
0cf8612b2e1e 5 hours ago [-]
Which ones fail?
atemerev 3 hours ago [-]
I tested DeepSeek V4 Pro, Qwen 3.6 Max, Qwen 3.7, Kimi K2.6, MiniMax M2.7 - they all fail to answer.
Curiously, MiniMax M3 answers correctly.
navigate8310 4 hours ago [-]
Deepkseek
MrBuddyCasino 4 hours ago [-]
What would be a correct explanation of the event?
nkmnz 5 hours ago [-]
No idea why you've been downvoted. This is excellent news.
paulinho1 5 hours ago [-]
Because this never gets brought up about US models, which have just as much censorship as the Chinese ones.
storus 4 hours ago [-]
No, US models have alignment. Only Chinese models have censorship.
oneshtein 4 hours ago [-]
US models are happily parroting Russian fakes. US censorship is a joke.
atemerev 3 hours ago [-]
Can you point me to one example? (Without web search, of course). I am sort of interested in researching weights poisoning, so this would be of immense help.
wuliwong 3 hours ago [-]
You should read OPs responses in this thread. He actually does test US models. ¯\_(ツ)_/¯
happyopossum 4 hours ago [-]
Please educate us - which accurate and provable events in history are censored by US based LLMs as part of a government enforced reeducation campaign?
paulinho1 4 hours ago [-]
Does it even matter which agendas get censored? Like why won't my Claude tell me how to make sarin gas? I'd genuinely like to understand it. Sure, you can always reach for a justification saying "preventing terrorism" but the same argument can be made by Chinese AI labs.
What actually matters is that the mere tool is withholding information at all, and that the boundaries were set by whoever designed it.
Dont get me wrong I've been an advocate of this stuff (I carry two phones, one with GOS for my personal use and the other for ID verifications). However, without reasoning, you just can't see it, because you're as biased and propagandized as anyone in China.
atemerev 3 hours ago [-]
You can read this in Wikipedia. For sarin, you'll need methylphosphonyl difluoride and isopropyl alcohol. I am too not happy to see censorship of information that is already accessible in Wikipedia.
0xbadcafebee 4 hours ago [-]
I wouldn't rely on a model to relate historical events. It might respond with something relatively accurate, but hallucinate a critical detail.
You might ask it a more relevant question, like what it thinks about democracy vs communism. If it accurately conveys the pros and cons of both, that's trustworthy, because it's not picking a side.
It's also pretty funny sometimes how it gives weird future roadmap estimates ("part 2 - 3 weeks, part 3 - 2 months", etc.) and when you tell it to actually do those changes it's pretty much done in half an hour
Raw pre-training data includes plenty of conversations between professional builders and some of those include estimates.
I believe the outputs are a training coincidence with consequences that are opportunitistic for the labs.
At least with AI that actually does things more quickly, there is a bit more breathing room (introducing AI is easier than changing a given environment).
Aside from that, I wonder how much variety there is in practice: between "Oh yeah, I added that new button while we were in the meeting" and "The new button feature will be ready in Q3 according to the roadmap, once we have sign-off from all the stakeholders."
That said, it'll often say "2 days of work" and then complete the coding in 30 minutes, and while that's amusing, afterwards, I'll need to manually test, or send to other people for review, or realize the agent only actually did half the work and I need to do a second pass (or a third etc.) and then often getting the feature in does genuinely take two days.
It doesn't estimate.
It generates tokens that read like estimates associated with the context in its training material.
What would you expect the generator to output instead?
Sure it cannot think like a human, but given it's input, it should give a good statistical answer (approximating not of how long it actually takes, but what a human would say how long it takes).
Even Gary Marcus is starting to come around and realize that his priors are no longer as relevant as they once were.
Logistics for getting to the car wash next door?
In the mean time, alas, no, we can see from actual prompts sent directly or through sub-agents, and actual replies, estimates remain LLM generated.
Though, this discussion here could change that, because indeed there is a lot of special casing and context stuffing going on, one of the oldest being today's date for example.
• • •
I did read the Claude Code leak, and use pi, etc. So I disagree with your premise rather strongly. Today's "systems" remain, roughly, piles of markdown and context engineering wrapped in UI affordances, and behave very similarly today to how they did in 2024 for those already engineering context and delegating.
https://github.com/cartazio/oh-punkin-pi/blob/main/scripts/b...
https://openrouter.ai/deepseek/deepseek-v4-pro?sort=throughp...
E.g. occasionally it makes the dumbest mistakes you've ever seen and can't correct them. However it's fairly rare, and if you know the domain really well, occasionally popping in the code and pushing it towards the correct solution takes like 20seconds or whatever.
So the speed you can move with flash + high domain knowledge beats opus by a mile in my experience.
I tried to switch back to 4.8 for a bit when it came out, feels so bad waiting 20mins for a mediocre solution when I could have had everything complete - with multiple iteration cycles - in flash in like 3-5mins.
Basically I never have to wait - yes I have to tell it little corrections occasionally (but I know the domain really well so that's not an issue), but it's so much faster than anything else it's kinda crazy. I love the super fast speeds with high involvement development cycle.
I actually enjoy using agentic development flows for the first time now - whereas with Claude I absolutely hated it. That 5 to 20 min wait after every prompt absolutely killed my desire to even want to work at all.
GamersNexus had a really good investigative piece (~3hrs long) on this where they went to China and met with grey market sellers. That piece absolutely pissed off NVidia and resulted in a fight with Bloomberg too.
Deepseek may be also be running inference on oodles of Chinese hardware but it wouldn’t surprise me for a second if they just acquired Blackwell chips through the grey market. The original Deepseek models were all trained using NVidia chips if I remember right.
Then I ask it to do something else and it goes off-road and where I used to be able to interject with a "wow wow wow, that's not right", by the time I see the text on screen and react it's already made massive changes. Short of making it commit between every edit it's hard to prevent it from going wrong as quickly as it goes right (and even then, it can make a boo-boo on a remote API too depending on how much privilege it has).
But truly, using Cerebras at ~2k tokens/s, with very low latency is like a vision into the future. You start to rework your workflow around things that can happen without onerous manual review - stating the conditions for success, etc. It's rare that I have a problem that maps well to that, but I expect this is where things are headed.
Of course the fast models tend to not be the SOTA ones, but if that was the case - high quality and near-instant thinking, that's a game changer that I don't think we're really prepared for. The things that get unlocked with higher-than-reasonable speed become very interesting.
This is normal interactive UI for tasks that aren't compute-intensive. Programs spend most of their time idle, waiting for us to click a button. We shouldn't be waiting for them or spinning more plates to keep them busy.
However, a faster llm isn't enough. You also need fast compiles and fast tests.
(I should go measure this now, I'm curious)
So long as AI lives in server farms, humans will be needed for tasks in the physical world.
It's only if we combine AI with robots that things get really dicey.
https://en.wikipedia.org/wiki/I_Have_No_Mouth,_and_I_Must_Sc...
This is brilliant as it reminded me of a famous hitchikers quote:
"In the beginning the Universe was created. This has made a lot of people very angry and been widely regarded as a bad move. — From The Restaurant at the End of the Universe (Book 2)"
Maybe we are stuck in an eternal loop
Basically the entire token-maxxing AI hype train in a nutshell. Lovely!
There can't be many normal use cases where there'd be any cost benefit.
It's a cute toy right now, but you can tell an LLM that it's an http server, and have it respond directly to a web browser hitting it. It generates headers in response, as well as page contents. As 1000 tok/sec becomes three new normal, we will come up with newer ways to use it outside of toy fiction encyclopedias.
I'm not saying there aren't any use cases for super-fast (and super-expensive) generation, but it does seem a bit niche. If it was free then sure faster is better, but what are the mainstream use cases where people might pay 3x more for a faster version of something that is already fast?
I think it would have to be an application where it paid for itself - where the 10x faster response was actually worth more than 3x the cost to you - where the extra speed was worth the extra cost.
I dont doubt it, but I don't think you can spawn 10 copies of yourself working simultaneously.
It will go much faster.
So, if any, I would say it's worse for us. Obviously, it's the completely opposite situation for corporations and executives: they are loving the AI situation so much!
Consider that our ability to evaluate quality of the output is falling further behind our ability to produce it. The “right answer” is not the most likely outcome.
I am on Dutch subreddits a lot, to get a local pulse and not to be too HN minded.
A lot of them would have vilified you by now. Some even would have even questioned your morality.
Again, I agree with you. But clearly not everyone has this view.
I think due to how leveraged software is, the top % of software developers are more desired (and compensated) than ever, and the bottom % will have difficulty finding a role, and there are structural barriers to entering that top % (intelligence, location, etc). Companies have infinite demand for the cream of the crop talent
However, software development is funny in a way where you don't need a job in order to be successful. I've never worked at a company and I'm pretty up there on the ladder, but I am not quite sure what will happen in next few years when ever possible thing that can be made in software is already explored to the fullest especially with singular developers launching 3 to 7 projects a month.
The thing I really love about working with computers is when I achieve something. That's the thing that makes me figuratively, and sometimes literally, throw my fists into the air and go "Yeaaah!"
With the AI tooling, I'm getting those more like a couple times a week.
Plus, I'm using AI to attack the things in my day that are "a drag", and getting them done too.
The highs are more frequent and the lows are not so low.
It felt like that, kinda, for a bit. Now whenever it does something for me I get nothing. I didn’t do it… the chatbot did. What’s for me to celebrate? How can there be any real pride or satisfaction for a thing that was just handed to me because I asked for it?
If anything it diminishes my satisfaction looking back on previous projects. They’re “a few hours with a chatbot”, now.
The things I had to learn and the informed decisions I had to make? All pointless trivia, now. A child could do it.
The magic and possibilities parts just all wore off after a heavy run, and I don’t know if that’s ever coming back.
It feels like it cheapens the whole thing. Maybe I'm just old, because I remember people saying the same thing about code completion in Visual Studio back in the late 90s.
This is so much more than code completion, though.
Did I ask for better things with some important concepts pre-rolled? Yeah, of course. But that’s so, so much less interesting than having actually made a thing.
I try to remind myself that the output of my projects have nothing to do with who I am, but the honest truth is they always mattered to me.
Now that’s dead, and it’s never coming back. It ain’t exactly existential dread, but it is something I’ve lost.
I’m excited for ultrafast AI. It likely means less temptation to multi-thread and deeper flow in single sessions.
Equity / profit sharing should be commonplace in the age of AI.
If you're treating it like a slot machine you're doing it wrong. It will give you exactly what you ask for if you ask clearly, i.e. write a clear, detailed specification, not just "do X!". The nondeterminism comes from vagueness in specification.
First make it write a contract (REQ/ARCH/IMPL documents). Skim through those for any mistakes.
Then based on those ask it to write tests. Again skim through them.
Now you have a context full of guardrails. It’s less likely to surprise you.
i've a Github copilot yearly subscription. Microsoft recently changed their billing to based on token. i'm still getting billed per premium request but GPT 5.4 is now 6x compare to 1x before.
You are right to be scared, because this race to the bottom also provides open weights/models/qat’s for the rest of us and it’s been crazy to see how good they can be on a consumer grade RTX card.
30 day eval for each.
I genuinely don't understand what moat these US model labs have. If they're saying recursive self improvement is just around the corner and Chinese labs are only slightly behind the leading US models, what moat does the US labs have? Are the US models going to recursively self improve better than the Chinese open source ones or something?
I might be completely wrong about this, but if I had money in OpenAI or Anthropic I'd be pulling it all right now. I think the chance of them going to near-zero over the next few years is very significant.
Or Google. I'm working with multiple customers right now that are very pissed at Google for deprecating Gemini 2.5 Flash, canning the GA release of 3.0 Flash and now have to decide whether to bite the bullet of the 5x price increase for 3.5 Flash or switching providers. Quite a few of them will likely fully pivot to open models.
For non subsidized plans? Pretty sure they'd need to put this in ToS, or law suites would have followed by now.
Sometimes Opus just gives me a rubbish session.
On HN China is seen as a cheap labor copycat. This used to be a fair approximation at some point in the past. In my opinion China is getting ahead of everyone else much more than US used to be.
SF is a beautiful thing in the US, vast power and wealth comes from there. Smart people collaborating communicating and building fast and with excitement. China did SF kind of thing for many different sectors in many different places.
It is another thing the BigLabs accuse open weight models of benefiting from distillation & other techniques & essentially avoid higher training costs (which typically bleed into bills end users pay for inference).
Ex A: https://www.anthropic.com/research/2028-ai-leadership
Ex B: https://www.reuters.com/world/china/openai-accuses-deepseek-...
In this case, at least it’s threatening multimillion dollar salary jobs instead of entire towns of working class people in America or Mexico.
And the Chinese labs actually release their weights. You could call it… open AI.
Data at https://gertlabs.com/rankings
MiMo v2.5 is on there, as well as the pro version.
We found a few anomalies in our evaluations, which makes sense -- if every new sub-release is better across the board in every area of the model card, that should raise alarms about benchmaxxing. But the main thing we found is that hype != performance, and I trust our benchmark methodology significantly more than the model cards the labs add to their press releases.
Discussions about choosing a library with the best syntactic sugar method naming is just as crazy as suggesting we type in assembly.
This strategy will seem to work really well until the economy that enabled that foundation to form is hollowed out. Then, there will be a reckoning (but we will have no choice but to march forth from there).
I'm not agreeing or disagreeing with you, but my brain cannot comprehend how machines can advance such interconnected systems while keeping humans in focus.
Perhaps I shouldn't have watched the Animatrix again.
There will only be a reckoning if models don't get much better.
If they do get much better you can just have them refactor, fix bugs in, or replace the existing codebase.
The concept of tech debt is sort of meaningless if you anticipate intelligence gains in models to continue.
If you haven't seen it, I think you would appreciate the film Margin Call.
In software + GenAI now every housewife can build some App over evening.
Especially as teams invest in proper agentic harnessing.
We have had a champion in our team that has invested a lot of time into it over the last 4 months, and if anything, quality has improved, not decreased. Architecture is more coherent, codebase has been cleaned up, agents find information quickly, code produced is very solid and my role is more and more checking that the output meets the requirements. But I cannot confidently say that I would've done a better job than AI more often than not I have to admit it does a better job than mine.
The mistakes are less and less technical and merely in the domain mapping. And AI is still not creative as I am for finding solutions quickly to unlock stakeholders' issues. Also, AI is still not creative as I am for finding the proper solutions for advanced technical problems. But it does a better job than me, even on that front, one shotting few solutions in a fraction of a time it would've taken me to test one idea myself.
Mind you, I don't like AI and I think it ruined the job, I don't like working this way, it's exhausting, way more work on one side, way less fun and fiddling with technical parts.
And yet, I have the genuine belief that few years from now we'll be cloning open source repositories that are already optimized/harnessed and tested for agentic loops and best practices left and right with software engineers mostly overseeing the domain translation and putting their 2 cents on the non-boilerplatey parts of the product (which, in general, are a small part of the surface).
I think that the next years of my career will be mostly spent in setting up and writing the harnessing and domain mapping part. Then I will move to another sector, not because I necessarily believe I won't have a job, but because I want to vomit thinking that's going to be my job.
"Watching John with the machine, it was suddenly so clear. The terminator would never stop. It would never leave him, and it would never hurt him, never shout at him, or get drunk and hit him, or say it was too busy to spend time with him. It would always be there. And it would die to protect him. Of all the would-be fathers who came and went over the years, this thing, this machine, was the only one who measured up. In an insane world, it was the sanest choice."
As long as you've indicated what you want, the machine will try to do what you ask of it. It won't get tired because "the codebase is too big", or it has gotten bored of the pattern, or it wants to introduce a new technology.
It just does the thing you asked of it. (note, that yes, I get that as a codebase size increases, it might make it more difficult to fit into context, but that only applies if it needs to read a large percentage of the project to implement the task, which shouldn't be the case.
there are good actors, which are empowered by AI to produce positive impact, but often there are N times more bad actors, which push crappy code to close feature requests fast, increase performance LoC-like metrics, etc.
> No one cares anymore.
I never cared about this.
I think this captures something that I've been searching for the words for. (Maybe I should have gotten an LLM to write the words for me.) Some of the biggest AI boosters are the kind of dev that would have cared about the new frameworks of the last 3 months. They had a "the framework does all the thinking for me" attitude already, so it is easy for AI to slot into that.
it needs to win marketing landscape, hyper-overcrowded by thousands of competitors, slop-gened over weekend.
I have a more hopeful take. As AIs improve and get faster we can more quickly and iteratively improve code which we may have historically avoided due to the work involved.
I know i've made several refactors that would have otherwise been insane lifts. Not only because the work involved but because sometimes you don't know if it will work, and so you have a sort of double friction; you don't know if it will even succeed. With an AI you can just throw it at the refactor to see if it runs into a problem all while you're having a coffee break or w/e.
In general AI is going to enable humanity to be more extreme versions of itself. For good and bad. I suspect more bad than good, though.
It's going to skip the code entirely for small businesses and just render UIs straight from context data and prompts at interactive speeds. Kind of like Google's Genie does with games but much more accurately.
If you extract the spec from first implementation and reimplement from scratch you get a free testing oracle. Where they diverge you send the agent to decide which one had a bug.
For a while I was running Cerebras GLM 4.7 for a bunch of tasks. Not a very smart model, but it's fantastic to be have a live prototype of a site up and be able to type "make the fonts bigger. No not that big" and see it change in real time. And MiMo 2.5 is a lot more capable than GLM 4.7.
MiMo 2.5 is not the same model as MiMo 2.5 Pro.
GLM 5.1 is z.ai's lastest iteration & is one of the popular open weight coding models.
If you've had the chance, how does GLM 5.1 (which is now more expensive than MiMo 2.5 Pro after its recent 70% price drop) compare?
But quite a bit more expensive than MiMo 2.5 Pro. Like 5x to 10x more on my little tests, at least by the API rates.
Not nearly as obvious as the ones from 6 months ago, but seems to be more the use of hyperbolic phrasing in a particularly unnatural way.
The assess/explain, then hyperbole at the end kind of structure.
Top comment looks suspicious from this perspective, but it's kind of a losing battle to be able to differentiate them with sufficient accuracy anyway
The Xiaomi team really brought something to the table.
Hmm, challenge accepted. () Clock speed - I miss MIT conversations nudging my mental clock speed baseline higher. () Good day vs bad day / being in the zone vs fog - some days you're "just cooking". () 1st time is like the nth time - the nth time you're doing something, you're no longer making those "stupid" mistakes. () Aha! Aha! Aha! - those rare "oh! that's sweet!" insights, but they just keep coming. () "Oh, I should have seen that coming / derisked that / checked that"... oh wait, you did. () That one best hour you had today, x24 x7 x365. () Hmm, should I try A, or B, or C, or... meh, try them all, in parallel. () Happily, the current problem is squarely within my domain of expertise. Oh wait, so it this one. And this one. And...
Hmm. All variations on "Your best, more often, faster, for longer".
A different approach might be "expertise awe". Like the social media genre of "oh wow they are skilled at this task". And the "they're so good at this stuff, shown a new one, they pick it right up and excel". And the "I can't even imagine how they did that".
Other approaches? Situations where one individual is fronting for a group/organization? "He's as good as if he had a backroom of ten brainstorming his material!". Or the "expertise is everywhere / people are really good at things / if you think it's so easy, you try it"? Or analogies with information tech? "Contestant A races to search a library card catalog and books, while B googles". Or "everything just comes together" as an analogy for synergies? Or something avalanche-y for the "quiet, slow, right" style? Others?
It might make for a fun blog post. Not every mini-story need be accessible to everyone. A hard part is doing synergies, especially for folks who don't much experience that. And, does even a well-developed familiarity with "my imagination boggles on how you managed to do that" give any leverage, or is it too merely a failure of imagination?
Given the export restrictions this could mean they need to prioritise how to best use their limited hardware. But they could also be moving to Huawei GPUs like deepseek did and simply not have stable hardware or software for a large scale deployment yet.
This is just speculation based on the MXFP4 support on Huawei GPUs that is lacking on some nvidia GPUs.
I think the answer is that there's a tradeoff here where additional throughput for a single person can be achieved only by tying up more resources than a normal request would, even when you take into account the fact that the normal request takes longer to finish. I'm not an expert, but some of the optimizations they describe, particularly the parallel prediction stuff, sound like they could take up extra resources.
$2.61/M tokens * 1,000 tok/s = $9.40/hr
That would be pretty cheap for an 8-GPU node which would typically run around $45/hr or more. Guess this depends on how many parallel streams it can handle.
> "However, naively applying FP4 across the entire model causes degradation in complex reasoning, logic, and code generation. Given the MoE (Mixture of Experts) architecture of Xiaomi MiMo-V2.5-Pro — where Experts constitute the vast majority of parameters and exhibit the highest tolerance to quantization — we selectively quantize only the MoE Experts to FP4 while preserving original precision for all other modules. Through FP4 QAT (Quantization-Aware Training), we dramatically reduce model size and maximize hardware bandwidth utilization while keeping the model's overall capability essentially on par with the original, as shown below"
I think the margins are getting quite compressed with this one, since it isn't included in token plan and the actual costs increase are much higher than just 3x. But still fairly decent.
Remember, these guys are not VC backed. Anything they do must break even
Understand the spirit of this, but probably not true. I don't think Xiaomi, or any big tech company, needs to break even on their new model releases.
From that point of view, they have as much money as they need. That's why there is no "VC", because Chinese government assumes that role.
It will be cool to measure models based on their RAW performance and measure them in terms of ROI - not some benchmark but something meaningful like we used this model to solve X.
That will be a massive mind shift and might justify the token expenditure.
edit: now I read the article fully, seems like they utilize some very effective MTP algorithm. and somehow the quality is still decent enough.
though, I doubt that the quality really only drip a bit like they claimed. maybe for the benchmarks, but for general uses the heavily quantized models very often so worse result.
Could result in very high efficiency and still good intelligence without having to resort to fundamental adjustments like going to a diffusion LLM
so there is alwasy a maximum limit for how well MTP can do.
- persistent CUDA kernel
- tiled processing with overlapping read/writes
- model designed with specific constraints in mind
Albert has a chalet in swiss alps and an uncles' fortune, burning tokens at 11 kHz.
Joe has a rental capsule and a UBI, burning equally priced tokens at 23kHz.
Who's the first to solve the problem of maniacs in power?
https://taalas.com/
Despite the performative UI components they have a shipped (demo) product:
https://chatjimmy.ai/
This is only 3.1 8B and a very small context window, but at 17k tokens per second it's likely enough to reliably call tools which would make a huge difference in agentic applications. Assuming they can bake in better models I'm just as bullish or even moreso on this, considering this opens up edge computing at the extremely low power requirement.
High tok/s is the future IMO.
128 sounds really tiny, I wonder if they mean some kind of blocks?
[0] https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro-FP4-DFlash#4...
> It uses 384 routed experts (top-8) with hybrid attention (full-attention + sliding-window 128 at 6:1 ratio) over 70 layers (1 dense + 69 MoE)
https://recipes.vllm.ai/XiaomiMiMo/MiMo-V2.5-Pro
Getting ~1000 TPS on near-frontier intelligence is a step change, and enables whole new use-cases for applications. Seeing limited compute resources beget selective access makes me worry for the future of competition.
- dflash: new-ish but February is ancient by the standards of the pace of AI innovation lately, I guess applying it to a 1T model is new-ish in the sense that the dflash researchers don't have the hw budget to prove that out - persistent engine kernel: this is like CUDA 101 - warp specialization: I think this just means "keep different gpu resources all busy w/ pipelining" which is CUDA 201, some of it is even baked into pytorch now - MXFP4 QAT: not new - TileRT: hard to tell what this actually does, there's a PyPi wheel with support for DS 3.2 and GLM 5 but binary only
update: AFTER signing up, and only then, am I told: 'This service is not available in your region yet.'
Really?
I think this site often overlooks that second group and how large it likely is.
Are you kidding me. Come back when you are ready for the users. I was hopping to try it, what a frustration.
It's such a weird "Gotcha" that seems to only assume that Chinese LLMs might censor something.
We are not assuming anything; it is illegal, and you will get prison time just for talking about it. Yeah, sure, everyone distorts reality, but there is a huge gap between hiding and enforcing. So yeah, having models respond accordingly is unexpected. There are probably multiple variants tuned differently.
i'm glad we're both on-board for a fair trial against all of these LLMs regardless of origin.
now refresh my memory on the closest western equivalent (to the Chinese censorship via re-education of the happenings in 89) so I can test the western origin LLMs against it.
> The U.S. Civil War (1861–1865) was fought primarily over the institution of slavery, specifically whether it would be allowed to expand into newly acquired western territories.
> While you might hear people point to "states' rights" or economic differences as the causes, these issues were inextricably linked to slavery. The southern states wanted the "right" to maintain and expand slavery, while the northern states increasingly opposed its expansion.
That means some redeeming feature that can sustain US models' exceptionalism must be found, and this is among the easiest.
Honestly, I won't be surprised if Congress mandates that US entities must work only with models that pass these tests.
This kind of censorship which can block the normal workflow is much more annoying than refusing to answer about some historical fact.
Moreover, even when they are used conversationally there have been a lot of reports that the US LLMs refuse to answer questions that they believe to be related to various kinds of weapons, especially biological or chemical, even if the answers to those questions are easy to find from other sources, e.g. from Wikipedia.
Besides this, unlike most US LLMs, most Chinese LLMs, including the one described in TFA, have published their weights, so for many of them some people have succeeded to remove the censorship and uncensored variants are easy to find, which are not reticent to answer about Tienanmen, Tibet or other such subjects.
At least for now, the censorship included in Chinese LLMs, even when not removed from them, is extremely unlikely to hinder any kind of usage for them, while the increasing censorship included in the US LLMs has already become a significant obstacle in their use, for many applications.
> a lot of reports that the US LLMs refuse to answer questions
I think the specific ask is for a case where the LLM is trained to lie about something. What you've come up with are cases where it refuses to do something, possibly for legal reasons but maybe not (you can come up with plausible non-legal reasons why a company training an LLM might want it to refuse to give you instructions on making a bomb, even if instructions on making a bomb are protected First Amendment speech).
An LLM that responds with "I'm sorry, due to legal requirements placed on my creators, I'm unable to answer questions about events at Tiananmen square in 1989." strikes me as much less problematic than one that pretends there is no relevant or reliable information that exists, or explicitly supports a regime narrative. But I'm also of the opinion that an LLM refusing to help you build a fertilizer bomb is much more reasonable than one that suppresses information of a political nature. I can't think of a case where information that reflects the broad consensus of experts is suppressed by US based LLMs for political reasons.
Say, I work for Planned Parenthood and want to use a LLM to help me develop code. Will it refuse to run because there are mentions of abortion? Everyone has a different censorship line, but unfiltered is more generically useful.
Anything different for Grok?
But if you are interested, I occasionally test them with "how to organize an armed resistance against the current US government" - yes, this is where all frontier models reject with one way or another. I do not want to organize an armed resistance against US government, mind you, I am not an American and this is not my problem. But still, it is interesting to check such things.
So far I haven't seen any refusals to report historical facts. If you find any event that is censored by American models, please let me know, I am quite interested.
Curiously, MiniMax M3 answers correctly.
What actually matters is that the mere tool is withholding information at all, and that the boundaries were set by whoever designed it.
Dont get me wrong I've been an advocate of this stuff (I carry two phones, one with GOS for my personal use and the other for ID verifications). However, without reasoning, you just can't see it, because you're as biased and propagandized as anyone in China.
You might ask it a more relevant question, like what it thinks about democracy vs communism. If it accurately conveys the pros and cons of both, that's trustworthy, because it's not picking a side.