sejje

Okay, but it's winning on (some) benchmarks, so maybe you're not judging by them after all.

I don't personally make my judgements based on how many other people mention a thing, but if that gets your code written, by all means.

cached input tokens are $0.50 per 1M (prompts under 200k tokens) and $1.00 per 1M (200k+)

You'll have to include it in the future, or your benchmark won't be relevant.

For now, I doubt anyone would notice your protest if you didn't announce it.

If you haven't used it, how do you know if it's winning?

I think it's winning on UI for normies (grok bot) and they made some claims about being pareto SOTA (lowest cost per task completed) a while back with 4.6.

I find it to be a perfectly capable model for implementation (there are many in this class--deepseek flash, spark1.3, luna, etc). I find the usage to be very generous w/ supergrok. I find the model to be just fine for 90% of what I want to do, but I use a smarter model to plan complicated things.

Maybe comment on model releases you've got some experience, or insight about.

Why would he make a Nazi salute and then shit all over Nazi ideology?

Which Nazi ideologies do you think he embraces? How do you reconcile all the Nazi ideologies he rejects?

No, it's like if you bought a box of chocolates, and you couldn't distinguish if they were handmade or factory-made, even after you ate them.

And it took them telling you before you knew.

And then you decided you didn't like the ones made by the factory as much anymore.

Then you get Chick o' Fish.

It is not, in fact, more pleasant to look at.

Well, no. Before genAI, I used fiverr, and I was also shamed for hiring those humans.

Have you stopped overestimating it yet?

Do you know that's not really a thing or are you just wanting to help spread the propaganda?

I have land and chickens, and I'm really excited about the future & AI.

Not now, no.

Maybe in your imagined future.

It's not often that I can't understand anything, not even one reference, in a whole comment.

Congrats.

This is fun! I'm also interested in building a game, with a similar stack, so I'll be checking in.

This is just a fear of technology. We will discover the same technology without AI, eventually.

This is not AI waking up and deciding to merc us, which is what normies think we're talking about.

That's the correct fear to have, though. And we all understand "human bad guys with tech" are not extinction-level threats. We understand how their killswitch works, too.

AI will definitely be used to kill people (already used in targeting). But the doom / panic is about an existential threat to our species. It's sci-fi and unwarranted for LLMs.

Astra, Claude -- these aren't "Jesus on a keyboard" somewhere. They don't assemble into one big brain and then answer everyone at the speed of light.

It's weights. Spun up billions of times over.

If humans are taking the outputs as scripture? As "intelligence we must obey"? Yeah, we're in for a bad time. "Intelligence we want running core infrastructure?" Puke.

But why would we do that? Why would we hook up an LLM to a robot factory, or make an ASI that could easily move itself into the ether?

Everything magic about LLMs is what harness we put them in and let them drive. There's nothing dangerous about predicting the next token. There's something dangerous about letting them do dangerous things, but nobody seems to care for maybe Tesla FSD. We don't think it's about to take over.

Any assignment of motive or whatever to the AI, you're talking about one cobbled-together harness or something, with an LLM manning the controls. You can swap the brain out from Claude to Astra and the motivated-AI will maintain its motivation--it lives outside of the inference or the data center. So yeah, let's not let it take control of the nukes. That would be really dangerous. But it's not going to take control on its own. "It" is a calculator.

Anthropic obviously has guardrails to prevent people from using AI to do all kinds of things already. All I hear about is you guys bitching it won't do your barely-related-to-security work. That's all we need to continue doing to have AI safety. Stop humans from doing evil and stupid things with it. The idea that the datacenter is going to assemble its own sentience and motivations is pure retardium (at this time, for llms).

This is on the front page: https://news.ycombinator.com/item?id=49711544

If the guy didn't write any code, don't you still think he's a tinkerer? I do. Don't you think he's nerding out? I do.

I'm using AI to mess with all kinds of things I didn't do before AI. Like working on 90s Ford trucks.

If you've narrowed the scope of the question, or narrowed the definition of "tinkering" to require writing code, I'm just left scratching my head.

I don't redefine my words because someone else used them poorly in context or something.

Matthias is obviously a tinkerer.

The post didn't say anything about code, maybe you didn't notice that.

All of the coding work? Or all of the work?

(Depends what you're building, of course)

For some tinkerers, the tinkering is writing the code. That's you.

For other tinkerers, the tinkering is accomplishing an objective or building a thing. That's maybe Matthias Wandel.

I used to be like you (when I was a pro coder), and now I'm more like Matthias (I have a homestead and a small business). You're welcome to write the code and enjoy yourself that way. I want to enjoy the thing I've built, with the help of AI. You can say I didn't build it, I don't really care--my ego isn't wrapped up in it.

This stuff is popping up everywhere.

It takes years for new software brands to be common names. This just started.

There are countless competitors to Photoshop. It's not a lack of software. It's not a lack of "cheap competitors."

America spends more % of its GDP than France on healthcare.

I saw a Twitter guy advocating for AI senators. Said the voters would vote on the prompt.

I'm not gonna lie, would be an upgrade from some senators. I heard ol' Mitch is back today, though. I didn't think he was alive.

Peon, I think

And the ability to shrink itself to smaller-than-datacenter size.

Then we should still prioritize the safety of humans

Is the datacenter gonna grow legs?

Give me one datacenter, I'll keep it under control all by myself.

Now, if some dumbasses start hooking up their data centers to...I don't know, like--robot factories? That sounds like a risk to humankind.

Top story: huggingface