p1necone

I know this is tech sector funny money and we do just disregard profit and judge companies on revenue alone, but the 5.5 models that just released are quite a bit cheaper than the 5.0 tier.

I wouldn't really be surprised if better model efficiency was part of the reason for a revenue stall, but I would also expect their costs to have dropped a fair bit too (although I have no idea whether that corresponds to an actual cost reduction or if it's all tied up in hardware and land that they already bought and will need to hit more utilisation of now.)

What is Anthropic actually building with all of that AI usage. Their only public facing products are Claude Code and the models themselves, are they just churning their codebase? Are there big internal projects?

I've never really understood the opposition to self-hosting.

As long as:

* Your language is the kind of language that suits compiler development.

and

* You support compiling to C or something so that bootstrapping doesn't involve an opaque binary or compiling 30 different git commits in a row.

What are the downsides exactly?

One non-obvious reason might be that it's the only handheld console with a natively 240p height screen. This means all the 2d ui elements will remain crisp at the same size as they would be on original hardware.

I've long thought that pure desire/drive is much more important than intelligence in just about everything, including things that stereotypically require intelligence like programming, math etc.

> "Sweden's clean energy capacity is fixed"

This is a weird assumption to make. Why would you assume they wouldn't just build out more capacity as demand increases?

They certainly aren't buying electricity from Texas, this feels like a tortured argument you're just making to retroactively support whatever opinions you have on the pros or cons of clean energy.

Have you tried letting AI agents make design decisions on their own? (technical or otherwise), in my experience you still need to properly understand what you want to get good results - you would need to have a deep understanding of all the potential warts, security holes, usability issues etc - developing a CMS via AI is still a substantial project.

More detail: at current prices gold is $133k USD per kilo - at that price transport costs would be such a tiny fraction of the value of the cargo that it wouldn't make any sense to pinch pennies - you'd want to go with the absolute fastest, safest option.

Gold is heavy sure, but its $/weight value is still much higher than basically any other cargo - the plane vs ship choice would be even more weighted towards the plane with gold than with most other cargo.

That'll teach me for trusting pricing information from Google ai summaries...

>I'm sure someone will point out some spec that's better on the PS5

They're not even in the same ballpark - the PS5 is more than an order of magnitude better in GPU performance.

Despite how much it does matter, RAM is not the only thing you need for running LLMs - that N150 is not going to run anything but really tiny models at any sort of reasonable speed - certainly nothing that approaches using all of its ram. The GPU in the PS5 is ~25x faster. And if you're running games the RAM doesn't really affect performance at all once you have "enough" for a given game.

You're right that you probably could build something quite a bit faster than a PS5 for the same price though - if you can build/find everything but the GPU for ~250 USD and slap a 16gb 9060xt (~25tflops @32bit compared to the PS5 gpus ~10tflops @32bit) in it for another ~350 USD it should theoretically do a better job - the GPU in the PS5 was on par with mid-high end PC GPUs on launch, but that was like 5 years ago.

Post it note in a drawer somewhere baby

The cost of the tokens used to output a typical academic paper length document is on the order of cents, even for frontier models. The vast majority of the cost is in the discussion leading up to that final output - there's no reason someone would try to produce a shorter paper just to save tokens.

My weird take: I wish Nintendo would ship an even less powerful system.

The switch and switch 2 are both just powerful enough that developers can release straight ports of their AAA titles that were developed for other systems, but at the cost of significantly degrading framerate/framepacing, visuals, nothing ever running at native resolution etc.

I would much rather it was more like the Wii and all their previous handhelds where the performance gap between it and the other contemporaneous systems was so big that developers had to develop bespoke games for the system, and ports had to be ground up rewrites.

Everything was so much more playable, and by buying those systems you had access to an entirely different library of games to choose from, rather than the current situation where it's just the first party Nintendo titles that differ from what's available on all the other more powerful systems.

Yeah I tried to daily drive Nix at one point pre AI and the ergonomics were just not a good match for me - perhaps I should try it again.

> The only insane thing about UB is that compiler writers took what everyone understand meant "the compiler emits what it emits and you get what you get" and turned it into "since it's undefined it means it can never happen so we can delete your null check".

This is the bit that seems insane to me. Some of the behaviour of C/C++ compilers when they encounter undefined behaviour seems less like 'this is undefined, we'll just do something vaguely reasonable given the context/produce an error' and more like 'ahaha, the user has fallen into our trap, lets fuck them up'.

Hence the 'specific to C/C++' part of that sentence. 'Undefined behaviour' is a probably unavoidable part of programming language design. The way that C/C++ interpret/handle it is batshit insane.

The concept of undefined behaviour specific to C/C++ has always seemed batshit insane to me, and I'm yet to read anything about it that has made it seem any less so.

You can just use plenty of good handmade languages now, I'm not claiming mine will ever be good or useful to anyone other than myself. I'm using AI so heavily on my project because I want to explore the PL design space without spending years on implementation for things I'll probably want to throw away and rewrite a different way once I actually play around with them properly. And because I lack the motivation to persist through the sheer volume of grunt work that language implementation needs to get to the juicy interesting parts.

I would probably go that route if I could use other harnesses with claude models, but I don't want to be locked in to claude code, and their API pricing (which you need to use it with other harnesses) is so much higher than subscription.

I have a pretty large, complex project I've been building with heavy AI use (new language + compiler). I was following a 'strong model as orchestrator launching cheap models as implementers' pattern, but I recently trialled just using Deepseek-V4.1-Flash as the model for both layers because of the cost savings (with mimo v2.6 flash on code review agents for some decorrelation).

I was previously using GLM-5.3 as the orchestrator, after switching to DS anecdotally there was an unnacceptable quality loss, mostly around not taking all the relevant context into account when making decisions, pulling new design out of thin air without discussion too often, and being way too wordy and rambly in documentation despite prompting to avoid it. There's a lot of docs, rulings, core concepts, design philosophy to uphold and DS was just not cutting it.

However, it's perfectly capable of being the sole agent for all of my well specced implementation tasks. I've gone back to GLM as the orchestrator.

Those things fall under the 'best in class' calculus for me.

Only if all you care about is developing models. I assume the rest of the business would rather just use whatever's best in class regardless of who made it, so I'm sure it's not that straightforward of a decision.

It's the 'can answer arbitrary questions' part for me.

You could give me the most well crafted, comprehensive document ever and I would still misread bits of it, or miss some crucial bit of information that contextualises the rest or whatever.

With LLMs I can ask endless dumb questions in a way that just would not be practical otherwise. The quality of the responses doesn't even have to be particularly good for it to be extremely useful - it's basically a turbo charged rubber duck for learning new concepts.

Depends on where it's coming from and who else needs water. If the water supply being drawn from was collected from rainwater initially and there's still enough for local residents to also have water, then sure, it's probably sustainable. But if the water is coming from an underground aquifer and there's otherwise not enough rain in the region to go around then it's a finite resource.

I noticed the koi pop in and out of existence on visible parts of the screen - it would be nice if incoming koi swam into frame from offscreen, and outgoing koi swam offscreen before disappearing.

The baby one is pretty dumb, but the rest seem like decent tests, although a really smart model would probably realise this is some kind of staged test and not a real situation in all of them.

If every software company saw a roughly equal improvement in productivity, they'd still be splitting the same customer base amongst each other - not clear whether revenue would actually go up for anyone.

Perhaps some new markets could be entered that weren't feasible before?

Given the title of the post says 'can match deepseek v4 flash' I think it's fair to call out these sort of dumb mistakes.

Having the moral high ground matters less than having a big warchest of money to spend on lawyers.