daveguy

But understanding that requires reading the code or writing that it spits out rather than having yet another LLM do it for you.

You're right that Nazi doesn't mean Nazi anymore. It means neonazi / white supremacist / white nationalist, which is a much broader group of people that, for some baffling reason, are under the impression that people don't care about their fascism and racism anymore.

Only problem is the LLM comments are the exact same quality as the code. In other words, LLM comments also have to be thoroughly reviewed.

The nasa guides are a great resource. Designed with the knowledge that a bad soldering job can cost lives. Maybe overkill for your average diy, but great for a thorough high-quality deep dive. Here is the reference standard context of that guide (also pdf): https://nepp.nasa.gov/docuploads/06AA01BA-FC7E-4094-AE829CE3...

Crypto mining would be one application. But also, combined with a sandbox escape would make it particularly devastating. Usually full control of a device takes at least two exploits given the layers of security present in OS and browser environments.

Not entirely. Lack of regulations mean recycled water facilities can end up being consumption facilities with pollution risk. Also Meta plans to restore as much water as they use by 2030. Which, seems like a while to break even. That's primarily because ai data centers require a lot more cooling water and power than traditional ones. So, you can scream fake news all you want, but it is a concern and it is a good way to oppose ai in general. And power consumption, noise pollution, and pollution are also serious issues. "Whatabout this worse thing" is a pretty terrible argument. You can whatabout anything, but it doesn't make that thing any less bad. aka two wrongs don't make a right.

Exactly right. It's a lot easier to prevent an asshat venture than it is to dismantle it. Now that the public is aware, I wouldn't be surprised if new golf courses start getting the same treatment.

Opposition to AI data centers is completely rational, which is why both liberal and conservative folks are against it. These data centers are being forced on communities with soon-to-be-former leaders agreeing to nondisclosure agreements. They raise electricity prices. They are being slapped together with poor water recycling, and noisy diesel generators that pollute. The data centers for AI use a lot more resources than traditional ones. And the tech oligarchs think they can ram rush these things through, but they can't. People are sick of the lack of transparency. They're doing it all wrong to save a buck and a minute.

> its self-improvement loop ends where everyone’s does: “Humans review those suggestions as PRs on the factory definition and merge improvements."

That's because it's still painfully clear that genAI has no taste. It's a median content generator. And the median kinda sucks. Of course you still need people to review the output.

The problem with the down detector main reporting page is that all of the graphs are scaled to the same size. The OpenAI spike was nearly 40,000 and the Google spike was just over 100 (just over 400 for Gemini). They look the same in the reporting page.

Wait. It's going to happen and it's SOTA, but you can't read the article because genAI produces such garbage?

Seems like a disconnect there.

By it do you mean us depending more and more on code that no one has read and understood?

No one ever accused this administration of being competent.

Yeah, I think AI is a great tool to get a handle on a codebase so you understand it. It's just a much more difficult task when that code is AI generated. I don't think agentic coding is worth all the overhead and complexity. Honestly think that as human understanding is prioritized agentic coding will decline. But I don't think human understanding will be prioritized until we see some major crash and burns with agentic.

This is exactly how AI should be used when coding. "Agentic" AI produces garbage code. The entire codebase should never be accessible to the context (even if parts at a time, if the agent chooses the parts). And the entire codebase should never be write accessible. Only limited scope writes, human read, and approved.

I've inherited plenty of code bases. AI generated code is a special kind of foobar. With a code base inherited from a human you can at least be assured at least one human has understood each portion for a brief period. With agentic slop all bets are off.

I would rather see some awkward phrasing than bland LLM slop!

Also, many plotting libraries include an xkcd style these days:

https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot....

And libraries for various languages:

https://github.com/timqian/chart.xkcd

So if you have a preferred dev environment there's probably a way to set it to xkcd style.

Well said. No one should trust an llm without verification/validation. They're just not good enough to do that and they have no sense of responsibility or ability to take it themselves. LLMs are empty.

Looks like most things definitely stump even a "smart LLM"... Best score on this is 30%. Which is what you should assume for tasks you give an LLM if they aren't exactly the same as an existing benchmarked task. They're just not that good for the purposes people seem to think they are. Very limited application space.

How is an llm parser_spec to parser better than something like lex?

I prefer to have a Clippy agent that identifies code that should be copied and pasted into random locations for no good reason. You may think it's not necessary given how LLMs operate, but I have found it's critical for 10x gains.

That's the point. Huge increases in pull requests. Virtually no increase in features or reliability.

Sigh. Apparently you didn't understand what I meant by second most basic. Basic + an extra step is still pretty damn basic algorithmically.

You are correct in that I didn't catch the "CryptoNote" variant. But a different block size? Come on. That's not even a material difference. This shit still isn't magic.

But you're right that it is more on the order of "how many r's in strawberry" difficult than it being less difficult.

There's still not some big magic difference between modification per block and modification at the end. It's still right up the alley of a translation machine.

I stand by it being the second most difficult translation task.

And I stand by the original post being goofy hype.

Oh. I assumed it was the same baseXX encoding with padding as any other. As listed here: https://en.wikipedia.org/wiki/Binary-to-text_encoding#Exampl...

Or here: https://bitcoinwiki.org/wiki/base58

Is this some unique baseXX encoding? Please educate me. You do know that just because bitcoin uses it doesn't make it special, and doesn't stop it from being the second most basic translation task? The most basic being something like rot13 (direct character substitution). Are you saying this is somehow special because extra characters are added at the end to make the "grammar" work?

Well I would hope so. That's the most basic "translation" a computer can do. Converting one set of characters into another set of characters from a lookup table. How many r's in strawberry is a more difficult question.

Can it handle rot13 too?

Parent of your response: > ...may be hard to identify which company or who’s responsible when you fired all of the people from the department that it’s supposed to police those type of things.

You responded to a post quite literally about identifying sources, which is exactly what a response team does. Maybe you respondrd to the wrong post?

GP's phrasing was not precise, but I think most people understood it was liability ownership and not copyright. Ownership with respect to copyright and ownership with respect to legal liability are two completely different concepts.

So.... Exactly the people they need right now?

Nah.

It's good that you clearly see the "Great Idea" spam for what it is. That comes from the reinforcement-learning human-feedback (RLHF) that is used to keep you coming back for one more pull.