dcre

Yeah, I agree — not a big difference for me between the Opuses. I really just meant Opus 4.x.

Claude Code (or Codex, or OpenCode, or Pi, or Amp — whatever) can do this out of the box without any skills or special tools. The most important thing for making results like this easier to achieve (in any harness) is using the best current models. Right now that's Opus 4.8 and GPT-5.5.

They're only good at it because that's what they're good at? Come on.

They are metered. That's why their ARR went from $9B to $45B in 6 months.

I must have thought I wrote something that I didn't actually write in the previous comment — I can't figure out what "as I said" is supposed to be about.

In any case, maybe I was too subtle. I was talking about Mythos, a model that continues the trend, but which is not available to the public yet. The "overwhelming evidence" is the testimony of the people who have used it. The irrational skepticism was people who don't believe that testimony. In other words, we do know the future, because we know that model and others like it will come out soon.

And there's a whole lot more evidence than that!

I think we have quite good reason to expect more. As I said, we already know (caveat with your level of irrational skepticism toward the overwhelming evidence) that the best existing models are better than the ones publicly available.

Judging from the fact that the Opus 4.5 inflection point was not really anticipated, and we still don’t really know what threshold was crossed that suddenly made agentic coding accessible to so many more people, I think it’s safe to say we don’t know what the thresholds will be until they’re crossed. The fact that we don’t know exactly what they’ll be isn’t a good reason to think there won’t be any more.

1. Global IT spend is $6T per year

2. Where does this $5T number come from? If they make $4T in revenue over the next 5 years instead, what happens?

I don’t think “Usage has plateaued except for coding” is compatible with lab ARR at $80B and still growing exponentially.

https://www.theinformation.com/articles/anthropic-openais-sh...

Weird is far too generous. It’s a travesty of thinking.

It’s not deviating up and down. It’s deviating upward. It is necessarily going to wildly overstate the previous 12 months’ revenue while wildly understating the next 12 months’ revenue. There is no way to describe exponential growth in a single number that doesn’t do this. This is why adults with a brain look at the series.

A new Mythos checkpoint improves significantly on the previous one (and beats GPT-5.5-Cyber) on this benchmark.

Most of Bun’s code is already written by LLMs. If you feel that way, it’s already been too late for a while. Furthermore, we’re talking about a million line port done in a couple of days. The question of whether it’s worth the time looks extremely different if done by hand. It would take a year.

Thank you for the link.

While this is a fundamentally stupid story to begin with, it was at least reported somewhat better in other venues. The original report came from The Information, and at least this Yahoo Finance[0] writeup mentioned that. This article has very little content and no sourcing.

[0]: https://finance.yahoo.com/sectors/technology/articles/ubers-...

Is that supposed to be good?

Too annoying; didn’t read.

My question is why AGI is required for these companies to be viable, i.e., why these companies cannot be viable in the case where AGI is not achieved. A response about what happens when AGI is achieved does not address that.

I think it’s true that we were able to establish trust and produce good work without verifying every detail — what I’m suggesting is that signals of that kind were not a very important factor. And code smells still work!

It's already out of date because it makes no sense. If it's true that the superficial signals of quality were once somehow good enough to keep the entire economy on the rails (it's not true), surely you can have an LLM look at given piece of work and extract comparably useful signals of quality or effort.

This does not answer my question.

Why is AGI required to make the investments work out?

SOTA models on medium are probably still better than free or cheap models, but you should experiment.

I recommend people look at the actual study and think about how representative are the subjects, the tasks involved (SAT essay writing), and the way LLMs are being used.

https://arxiv.org/abs/2506.08872

To be concrete, this is taking a task in isolation that LLMs can do much better than humans (writing garbage essays) and using LLMs to do that task. In the real world, tasks have parts and they exist in a larger context. When we use LLMs for one part of a task, there are other things we're doing that the LLM is not helping with. If you compared people doing arithmetic by hand and with a calculator, you would also see very big differences in how active their brains are. But it's not anyone's job to add up numbers. Adding up numbers is a subtask of a subtask in someone's job.

He wasn’t predicting slop; he was describing mass culture, which already existed when he was writing.

LLMs can tell you exactly how to acquire the materials and manufacture the materials. They might even come up with novel formulations that rely on substances that are easier to get. There might be information about this stuff online but LLMs are much better than random idiots at adapting that information to their actual situation.

On top of LLMs reducing the cost/difficulty, the other reason biological and chemical weapons are such a worry is their asymmetric character — they are much much easier and cheaper to produce and deploy than they are to defend against.

Sure! I don't mean they're all good. I just mean that it can't be cognitive offloading itself that is the problem, but the particular character of it.

I've never seen an argument like this that, if true, wouldn't also apply to the cognitive offloading we do by relying on culture, by working with others, or working with the artifacts built by others.

It seems like all of your comments are like this. Consider stopping that!