skybrian

Okay, sorry about that. But how do we start? Maybe there are AI safety organizations that deserve our support?

If you mean high-trust societies like maybe Switzerland, I agree that it can work, but the US has lots of guns and doesn’t seem very polite, so how we’re doing it doesn’t seem to be working very well.

We do have lots of food safety regulation, which has more to do with selling food.

Populist doomers will tell you that nobody can be trusted, no global problems will ever be solved, nothing can be done, game over, don’t even try.

People did use global agreements and regulation to fix the ozone hole, though, so I think there’s a chance.

This is a common populist sentiment, but if you don’t trust anyone then nothing can be done. Is it just game over?

We do in fact have car and food safety laws. Regulation is normal.

Okay, you’re asserting that but I disagree. Why should anyone else be convinced? Why can’t we get the good uses without the harms? It doesn’t seem like an unavoidable tradeoff.

There’s a libertarian sentiment that that doesn’t sit well with “AI is harming people” sentiment. If AI has harmful uses, and I think anyone sensible would have to agree that it does, then giving everyone unrestricted AI is likely to make it worse.

It’s sort of like gun nuts arguing that more guns is the answer. I mean, ok, maybe you’re a responsible gun owner or AI user but relying on personal responsibility doesn’t fix systemic problems. There are bad people out there.

In the long run, we’re all dead anyway. (Unless something drastic changes.) That means how long it will take for robots to be practical matters. We don’t know how long it will be, but driverless cars have been coming for 15 years now and they’re still unavailable in most places.

People are more adaptable than horses and tech solutions usually require people to do some of the work. Dishwashers need loading and unloading. My robot vacuum still needs maintenance. Self-checkout means you do the scanning.

Amazon wants to get into drone delivery and there are cases where that might work, but the way news articles describe it, it’s not going to get a package onto my porch.

This seems to be about ending support for ChromeOS early, but they also have a new OS that might run on the same machines?

And assuming the new OS also runs Chrome, Chromebook users might not see that much of a difference.

The $20/month subscription is an excellent deal. I wonder what these whales are doing with all those tokens?

I mean, other than Yegge. He seems like a pretty extreme outlier?

So... what kind of Internet access do dots have? Which bulletin boards will they use to compare notes?

I get decent results telling it to use Luna subagents for implementing commits.

In case anyone finds it helpful, I asked ChatGPT to break down what they found by app:

https://chatgpt.com/s/t_6abbbee386bc8191a26717b4f1442657

And how would they leave? What countries take refugees from Gaza?

When the Onion is not the only website that publishes satire, what’s that a sign of?

You can find examples of deluded people everywhere. It would be more defensible to claim that some intellectuals are idiots.

Suppose you use LLMs in a more straightforward way, like asking coding agents to make specific changes rather than attempting to build a software factory?

That could be seen either as early adoption that’s overfitted to current capabilities or as late adoption of LLM’s more advanced capabilities.

To get the context that isn't in the code, maybe it would be better to ask for a review of the prompt?

I believe they'd also be cached by proxy.golang.org? But companies should probably run their own proxy.

Not really convinced. I'd rather it stayed on Github so it doesn't disappear. Particular for businesses whose priorities might change.

Word-policing isn't going to magically fix laws or even identify which laws need to be fixed. It won't change enforcement priorities. It won't make companies more or less likely to sue for damages.

I'm also not sure it even helps conceptually? If you're interested in the technical details, by all means discuss the details.

Software developers do sometimes make promises to their users, but I think these promises ought to be explicit rather than assumed. You can't simply assume a "duty of care" and expect that other people will understand them the same way you do.

(Or rather, you can, but you will likely be disappointed.)

The Language Server Protocol is designed for advanced text editors (IDE’s). Now that we rarely edit text by hand, I believe we could instead use source code viewers that don’t have to deal with unsaved or incomplete code.

If we can assume that all files are saved and each file has correct syntax (which can be enforced by tooling), a much simpler protocol could handle all the language-specific indexing needed to make it easy to read source code.

(I’m assuming that we still do read source code. I for one like reading source code and have little interest in ‘software factory’ techniques where you don’t look at at the code. I might not look at every commit, but I’ll still read it fairly often and tell the AI to fix anything that’s unclear.)

> our work shows that what models say about themselves is not a fact about them

It seems like should be obvious given that they can play multiple characters, but it’s good to have more confirmation.

Although, I do wonder to what extent these personas might become stable entities. Could personas become portable and spread like memes? It seems like that depends on the extent to which prompts can become portable, causing similar effects.

Too soon to say apparently? They got some laws passed.

All that is true and I suppose there could be tax policies to encourage land reclamation, but this is more likely to run into environmental concerns. It’s literally an edge case.

The comments here are terrible. I got a better idea of what’s going on by asking ChatGPT what this paper’s weaknesses are:

https://chatgpt.com/s/t_6ab7d694885481918083b8cbf0ba9040

In particular: sometimes they measure “churn”, which doesn’t show whether the results are better or worse on average. They sometimes only test with one random seed. There are multiple-comparison issues. And they’re not testing Anthropic’s algorithm.

Changing a random seed could either improve or degrade the output. In theory, better and worse outputs should be equally probable, depending on your luck.

It makes sense to leave, but it also seems like the overall sentiment would be more positive given the “it paid my rent for many years” part.

I'm hoping that by "gets it," he meant that if you start a discussion about the design, it doesn't misunderstand and immediately go off to do the work. Some models tend to do this.