Hacker News Reader

Tenokeon Mistral raises €3B

It's possible to be a waste of European money that could be better used elsewhere, though. I think the same about LeCun's company sucking up the little funding here.

Tenokeon Mistral raises €3B

> Europe absolutely needs a home-grown AI lab, especially with Pax Americana looking increasingly shaky.

The only problem (at least in the LLM space) is that you can do more in Europe by just getting the best Chinese Open weights model (do a finetune if you really want) and do more for cheaper than using Mistral.

Tenokeon Any Human Ever – One life, drawn at random from all who have ever lived

Yes, this is exactly what I expected (showing the anthropic principle in action), so I am disappointed they are not sampling correctly.

Tenokeon Samsung's Processing-in-Memory (PIM)

>there ought to be a tipping point beyond which local inference is good enough

There's no such ought really. Even at current levels you'd need like a 100x gain from here to approach current top proprietary models (probably a lot more for say Mythos or Mythos 2), and it's not like they are stoppng to improve. This is before we even account that you'd just be running 1 agent then, and not a swarm like you'd be able to in the cloud or that you can do only so much compression before you are losing out

Tenokeon Samsung's Processing-in-Memory (PIM)

There's been a ton of optimizations already, it hasn't remotely reduced demand even temporarily. More efficiency just makes the compute have even higher ROI per $ and watt spent.

Tenokeon Nvidia agrees to acquire Hugging Face for $13B

Too base cynicism. I'd be willing to bet you my $200 to your $100 that doesn't happen.

Tenokeon Don't Paste the AI, please

That's a good formalization of it. I will think you are a bit of a fool if you dont use AI but you guarantee you'll be a fool if you only use AI with no value added by yourself.

Tenokeon Claude Opus 5

Fable has more parameters. In practice it's not yet clear which one would be better for different usecases yet but they are more different than one being strictly better.

Tenokeon Flux 3 X Mimic: The Next Generation of Video-Action Models

Mistral is much worse in its respective field than Flux in their own so I hope not.

Tenokeon Flux 3

Flux 2 Dev Klein has practically been the best you could use on most commercial hardware so I really hope Flux 3 has a comparable updated open-weights model to it. if not it'd be a great loss to most hobbyists.

Tenokeon Startup founders urge U.S. government not to shut off Chinese open weight AI

Y Combinator the company doesnt particularly have to share the opinions of hackernews the public site.

Tenokeon OpenAI and Anthropic unite against open-weight AI risks to their bottom line

Because sadly, as much as I wish it wasn't the case, offense is easier, more impactful than defense.

Tenokeon OpenAI and Anthropic unite against open-weight AI risks to their bottom line

I am very pro open source models - I use them every single day.. But we obviously don't want everyone to have capabilities like and beyond what caused the huggingface incident in every domain, so it's not like it all comes from bottom line cynicism.

Tenokeon OpenAI and Hugging Face address security incident during model evaluation

There's ways to make sure env vars get only injected at runtime and arent easily accessible otherwise or to even make them inaccessible to the user your agent is running on, and for you to manually run the code with the right permissions when the keys actually need to be used. Almost nobody bothers doing it though.

Tenokeon OpenAI and Hugging Face address security incident during model evaluation

That's kind of insane. Natural that it's happened, sure, but insane. I know people don't like thinking of it like that, but things analogous to this can easily happen in various domains with today/tomorrow's models given access and a different task.

Tenokeon Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

It's also very possible that they know their big model underperforms chatgpt 5.6 and fable by too much, so they are focusing on what they can get wins in like speed instead.

Tenokeon Fable 5 vs. GPT-5.6 Sol on an NP-Hard Problem: Does /goal help?

I do that but it ends up putting so much extra crap there that it has the same drawbacks.

Tenokeon Fable 5 vs. GPT-5.6 Sol on an NP-Hard Problem: Does /goal help?

Claude seems to forget what you tell it in very long work sessions (things that take weeks to develop), no matter how many times you tell it which part is extra important. I dont use goal (I guess I should), but presumably it makes it actually remember the most important instruction. I believe this here is about shorter sessions where the issue doesn't crop up as much.

Tenokeon GPT-5.6

>The human brain manages to self-organize with only a fraction of the information that LLMs get trained on.

So? The question isnt can we get to ASI as efficiently as a brain, the question is can we get there, which we likely can. The inefficiencies can also be fixed after that.

>No matter how many trillions of dollars get thrown at the problem, they still don't learn like humans do.

Again, so? Humans are efficient but also bad at many things that transformers are already better at because of it. You are looking at the wrong thing if you think it needs to be like humans.

Tenokeon GPT-5.6

LLMs dont just use text for a while now. It's also not fully supervised for a while.

Tenokeon GPT-5.6

You can watch the whole Lex Friedman interview, it's on youtube. It's not out of context at all. He goes on about how LLMs will never be able to do things that they do trivially. And he has just doubled down for years.

Ive read and watched more of his interviews and lectures it seems, it feels like you just have a rosier idea of his views than the views he repeatedly presents.

Tenokeon GPT-5.6

https://youtube.com/shorts/zQTt8TkcyfU?is=09r7XDqz2w6-Pygu

You probably wont like the edit but I dont have the timestamp of the original on hand, you can find it.

Tenokeon GPT-5.6

>He's merely said they don't think

He said years ago even 'GPT 5000' couldnt do things that they ended up doing fine a month later, let alone by 5000. His later predictions are just moving that goal post including towards them not being able to do more general, harder problems of which Arc AGI is a counter-example.

Tenokeon GPT-5.6

His main anti-LLM predictions have been consistently either wrong or misleading.

There's many ways to skin a cat so you can probably do something with a JEPA approach as well, but I doubt he actually catches up to having agents on the level of where Anthropic/OpenAI will be at any point.

Tenokeon GPT-5.6

Is any of those comparisons about Pro vs non-Pro (Pro is only available in $100+ plans)? I am curious about that but I think Sol, Terra, Luna are different sizes of it without the Pro part, and I want to know how much worse do I have it on the $20 plan compared to if I upgrade.

Tenokeon Mistral's Robostral Navigate: a state of the art robotics navigation model

8B sounds tiny. Of course, that's enough to easily run on device which is nice, but surely the actual SOTA must be some much bigger model?

Tenokeon Fable turned reMarkable into Tom Riddle's diary from Harry Potter

If you are imagining that, you could imagine it with search doing the same 10 years ago, which would have more thoroughly prevented you from researching things.

Tenokeon AMD Ryzen AI Halo – $4k AI Dev Kit

>but you cannot really hold a "portfolio" of it in any sensible way.

xAI effectively did and lucked out to cover their losses and more with it.

Tenokeon The AI Superforecasters Are Here

1. The liquidity is not infinite to compound that easily.

2. The alpha dries up with more players, even in the year or whatever since that founder started.

Tenokeon The AI Superforecasters Are Here

He's been interested in forecasting since a long time before there was any money in it, and still hyped Metacalculus (free research project) and Manifold (play money) because he is interested in the subject. If it smells like an ad it is because your ad sense is faulty.