minraws

Nothing is an easy to use intuitive api with all the features and all the performance, it's (Dx11) the sweet spot of where I felt like it was easy to learn and easy to use without worrying about performance too much.

Sure if you are building the newer rendering pipelines DirectX 11 leaves a lot on the table but it brought a lot of nice things with it, Debug Runtime alone was worth moving to it, 99/100 times your app written in DirectX 11 even if not optimised properly will perform better than DirectX 9 in my experience.

This holds with OpenGL though the modern OpenGL 4.6 and stuff wasn't bad, but it's a lot of cruft still.

Either way there are no perfect APIs, a truly general purpose API that just gives you basics like Texture handling and io and data mapping would be enough today imho. Fully general purpose compute APIs are the solution with some additional tweaks to make graphics stuff convenient. IMHO

Though it's definitely much harder than plain and simple paint colors and textures on vertices & geometry uv approach of old.

That middle ground is often called DirectX 11 but sadly it's stuck to a specific platform.

I don't think people realize that there was no option of do we not log the forest, but imho the issue is any forest being cut down for data centers is a waste of sand and effort, make them in deserts or dry land. In the middle of oceans works as well, we are getting data centers in all sorts of weird and imho wrong places.

I find this issue small but the choice of these companies to pick the convenient location for them to build a data center concerning, and govts shouldn't be bending over. If these things can be built in space we can build them in places where no one lives and no living being is harmed or logged.. cough.

This does feel like the fix took so long to bake and was so obvious way too many people who are into PHP knew about it, I knew about eMudra's cert issue as soon as Chrome and Firefox pushed updates to disallow them for instance because of reasons, and it wasn't even fixed after it was all made public on several platforms, I think Apple still hasn't fixed it properly, the trust Store is still fucked not sure if Safari was separately patched, maybe there is an os update in the last few hours I missed.

This entire RCE race is always annoying.

how are the secrets decrypted, I would like some explanation so that I can have some understanding here, I am starting to feel like all these companies are just meme-ing with their security posture.

Like did someone vibe code too hard, how do env secrets leak for everyone...

they now do something called, Meta Family daily active people, similar thing but aggregate of all Meta stuff. very sus imho.

They beat revenue expectations, and earnings per share was ok, and they guided for even higher revenue next quarter above estimates ofc.

Somehow Meta is making money despite starting to loose DAUs now. So honestly I don't quite understand what's what, also most of the big income is from Ads and biggest Ads spenders are startups and companies heavily dependent on VCs and AI... So it's definitely somewhat circular, and Ads is the most recession affected business so a down turn in markets will cycle through it.

But ngl if all the earnings are this ok then hyperscalers stock might not crash this quarter, well then again I didn't think they would but a person could dream.

I am quite bullish on AI, but would prefer the fluff to get out of the way asap, otherwise don't want it to happen at a bad time, and too many people get caught pants down.

I am concerned for my self hosted servers, well only hope is tailscale should be entirely secure, and egress is filtered heavily... If that doesn't work my home lab is fucked.

Any suggestions folks, not even my terminally online self can handle all of these.

Well we all got fucked by this cert issue.

If US govt bans it, I will move my business out of US lmao. I say that as non us resident who has business there for ease of doing business...

This isn't true, especially not with the inference engine becoming public, now you can build your on models on the same or similar architecture and ship massive gains on top of these stacks.

I think people don't realize the Salesforce SAP Palantir would love to ship these models, and they have/are shipping them.

The threat of exfilteration of data via these models is high but it's nowhere near as high as the margins.

For instance if you could cut your ai input bill by over 90% would you do it? Can you put some of that extra margin say 10-20% to make it more secure.

I know of atleast 10 fortune 500 companies that are entirely planning to or have moved to Chinese models for a large portion of their work.

It's insane people in finance are delusional enough to think that won't happen.

Maybe Govt will try to de-risk with bans on using open models but I don't think that's going to work at scale, people will literally spawn labs to launder DeepSeek/GLM models into a US model by some novel training etc approach or something.

I still don't think that it's as much a kami-kaze rather it's an attempt to normalize the world around the Chinese stack.

If DeepInfer is supported by everyone, it will be much easier to export stuff aka hardware that only supports it for now or is optimized for it.

Assume china starts selling hyper optimized Asics for inference at some point.

This feels like a race to the bottom as soon as possible because they have much lower costs, if there is a Wipeout in AI Labs in US chinese labs are largely insulated, and they can hopefully buy Nvidia chips for cheap if that does happen.

It's not really if you mean to point to the bitter lesson, it was models/intelligence scales better with general/simple methods with more compute better than it scales with any hand-crafted human program/method.

Ala intelligence is a function of compute, there was no guarantee that you need to 10x compute for 10% better performance or something like that.

If you ask the labs they will say the scaling is fairly linear or atleast sub-quadratic, they definitely won't say it's exponential.

No one was thinking we need to 10x compute from here again for another 10-20% increase in performance.

It's only been since Opus 4.6 that the compute growth vs improvements are starting to feel this way.

For instance GPT-3 vs GPT-4 was practically a 10-100x jump for a 10-100x increase in compute, we went from broken sentences to fully formed essays writing, speech and so on.

GPT-4 vs GPT-5 was just as massive we went from no agentic capabilities to fully agentic, 10-100x for a 10-100x increase in compute usage.

GPT-5 vs GPT-6 was atleast a 10x (upto 100x) increase in compute, but the difference is not that clear sure people say that one is significantly better than the other, but GPT-5.5 was good enough compared to where we are today.

Even on benchmarks compared to GPT-5.2 Codex we have almost 2x'ed the performance on say artificial analysis (53 vs 29). This is absurd. 10-100x increase in compute for 2-5x performance improvement is a massive gap.

[1]: https://bitterlesson.ai/ [2]: https://artificialanalysis.ai/?models=gpt-6-astra%2Cgpt-6-lu...

These models exist because better models exist, these models can't exist without better models existing, you will never beat the frontier with smaller model sizes. Because all these models are trained again larger models, generally a teacher model of sorts.

Lots of synthetic data, distilled data etc.

The problem is if we can pack a Fable in Qwen 4 27B who will pay subscription prices?

What's the process for getting access and then becoming a customer, I would love some transparency on how these programs work if anyone from relevant companies would like to elaborate.

Can I as a non-american even get access, how does it work if I am based in US for a bit, what about sharing access? Do individuals also get in? How big does a company/startup need to be, is there a tiered solution, what are the adjusted costs? and how much of the data is shared if we do get in.

There are so many questions, and I have zero answers.

So you are saying paying 25K isn't enough we need to pay more.. oof I got it, I am too broke for this world.

I think it clearly points to where they expect themselves to be, Opus 5.5 and Astra or Sol 6.1 are models that shouldn't exist at those prices without the price competition.

And no new releases from Chinese labs likely means their models that they could compare against those models are still baking largely because they lack compute so severely they need to train models for far longer, or figure out training on Chinese hardware.

Either way it's clear open models a lagging by ~1-2 months compare to frontier, depending on what frontier means to a person. Honestly most new models are just not that much better than at edge cases, they still fail a lot for the kind of stuff I do, but those edge cases matter a lot for un-interrupted long horizon work.

I personally am starting to see a wall here in llm development but that might just be me, given that people are being forced to scale models up considerably for frontier performance.

It's scary thinking that modern models will likely be at best 2-3% better in benchmarks or only useful in 1-2 unique edge cases compared to base models for 100x cost increase in compute required, it's getting absurd.

I can't imagine how Fable models can retain any value once you have to go 100x larger, while 27-32B models start to catch up to the current frontier for coding or other narrow/focused tasks in a year or so.

I created a pelican worked wonderfully, better than old haiku model or even sonnet models, maybe slightly worse than new gen models.

But overall it's a pretty ok model, the issue with speed seems to be that ttft is still high but still not sure how high thinking burden would be, I hope it scales well will be using it a lot if it works well.

This might actually crush most other LLMs if it's price and performance actually scales well. It was as cheap as deepseek for a few tasks I tried.

Anyone know if there is a subscription for it? I might want to test it out properly without burning a few hundred dollars on it.

Can we split Github, paid/pro users, my orgs are payings north of 25k per year to Github, can we get better service please, even AWS assigns a proper support performance once you hit few tens of thousands in spend. This is just insane, that all of github starts failing at once, I would like to have better experience for when I am working compared to when I want to spam commits with ai to slop something.

This is not true, Claude and ChatGPT have sent traffic to sites I maintain without any user-agent, it might be they have stopped doing this.

Google has done various versions of this as well, Grok too.

It could be these bots were in some custom harness or something, but they literally sent requests in very weird ways, that humans won't be doing.

I have ended up putting my sites behind vpns or anubis and similar. This is beyond painful.

Banks can keep it in US and stay happy, they don't necessarily immediately need physical gold. So it's quite efficient, lots of govts do similar stuff.

Qwen 3.8 wasn't that nice in my experience Qwen series has lagged a lot but I think Qwen 4 would be better they cleaned up the architecture.

And we are likely to see many companies start building 27-32B MoE models custom built for use cases now that Rubin is going to make training costs much lower.

I expect training costs to get under 100k USD within a year for small models with another 100k maybe for RL.

I think just having all your own data from claude and stuff on your system that you can feed these models and have companies offer great base 27B models will be how it works.

By next year we will likely see GLM 5.3 level performance in Qwen 4 MoE models that can run on most higher end consumer systems.

I don't think there is a reason to care about any of it, just get a few dollars on a few openrouter or similar platforms and just use the models for now, wait for cheaper models that are good enough for you to become available for your hardware.

Sure you might never see frontier performance on consumer hardware but do you really need that though? Does anyone? And if you ever do just get a 30-day sub for that work, and cancel it.

As unsurprised as I am by this, and as much as I suspect a lot of people could have guessed it, having Claude confirm it is still fascinating, but.

Somehow, I find these results fascinating, but also quite sad, because we used a machine to solve it at all. I feel like some half drunk motivated graduate level humans could have figured it out too.

Then again, once you know the answer, perhaps you have an obligation to share it, keeping it to yourself might be unfair to everyone else.

Honestly I’m not entirely sure how I feel about all this. I would gladly welcome any groundbreaking work Claude might do to solve hard problems and move the world forward by decades or even centuries.

But these sorts of fun problems feel like things we should avoid pointing LLMs at. They remind me of those weird programming puzzles that are difficult to solve but incredibly fun to work through.

My concern is that if Claude ends up solving all the fun problems, will humans be left with nothing but the hard, headache-inducing ones? Tao had raised a similar concern, problems that graduate students would have been expected to solve and learn the process, are now things you can simply point an LLM at.

I’m still not sure how I will ever reconcile with this feeling, about any of this. I just find it genuinely sad that we might be automating away not only the difficult work, but also the joy of figuring things out for ourselves.

"As a person you are not saving the server." Sure I believe you, I have complete faith in your works and in other humans, we have never had people sacrifice well being of other people for a little bit of money. That just doesn't happen.

Jokes aside, I feel like we attribute too much power of choice to these models, they aren't decision machines they are approximation machines based on what's the best outcome as per "average human understanding". This by no means is the "average human" output, but the direction is always towards the mean of what people feeding these models their thoughts consider acceptable.

This models aren't truth systems they are mean reversion systems, you can push them towards your mean with agentic harness and prompts, but they aren't decision machines.

For instance if just have to never give the model the choice to be able to do any harm, then it doesn't matter if it's system decides to do something wrong.

This is the kinda pointless and largely meaningless but very fun to look at stuff that I enjoy. Sad that it's also written by AI but it gave me a good chuckle when I looked at some of the projects I was involved in.

Can we make it work on top of some kind of Git clone inside browser local storage to work around the GitHub api limits or something? I think I hit those.

can for the love of god we get some usage in cheaper subscriptions being Ultra tier only is annoying. I want to just test it out here, well will burn some money via api I guess.

I didn't mean to imply that if I did that was unintended, I meant to say it's something that's going in the wrong way directionally, starting with recent heavy censorship, and folks in China don't have to good way to speak out or change things.

Even countries with back sliding freedoms with good democratic structures have a fighting chance, China is just so weirdly removed from it that it's concerning when and if something does go wrong.

I mean in companies you can just buy more, and I think AI compute is going to get a lot cheaper in the next 4-5 years, photonics are just around the corner, assuming AI is even 1/10th as good as AGI they should help prove substantially useful for just pushing photonics based computing over a couple decades sooner.

I went to China recently it felt different, the recent govt stance seem a lot more against the idea and the ones before which was my point, China in the 90s was a very different country from the China of today, then again I haven't been there then.

Again it's hard to say where it's open or where it's closed, they seem to be back sliding in very weird ways. I am not part of the LGBTQ community so I can't say for certain how bad it is, so I won't comment further but UN & human rights watch has been citing increasing censoring of LGBT representation and open-ness in China in recent times.

The biggest concern with China is lack of democratic freedom means as long as it's not bad enough to revolt the country can back slide pretty easily on many metrics with complete silence, no protests etc.

[1]: https://www.hrw.org/news/2026/06/25/china-social-media-films...

Everyone loves money but socially and culturally China is a lot more closed compared to US, much more so than what people are used to atleast.

US was arguably the most open country on the planet in terms of acceptance of ideas and cultural diversity especially in 20th century.

But I also don't think people need to be very worried but still we haven't had monocultures on the top or close to the top of power hierarchy for several decades so it's going to be uncomfortable for a lot of people.

For instance China is heavily censorious of LGBTQ and very likely against the Idea they haven't banned it outright yet thankfully. But then again lots of countries are in a similar boat.

But the biggest difference is China can easily back slide without possiblity of return on any metric because it's not a democracy.