dcre

Incredibly ironic to make this argument using such an abstract and low-dimensional framework.

I just don’t think the question is about determinism and probability at all. When we think, our thoughts are influenced by any number of extra-logical factors, factors that operate on a level of abstraction totally alien to the logical content of thought. Things like chemical reactions in our brains or whether the sun is out or whether some sound distracts us or a smell reminds us of some memory. Whether these factors are deterministic or probabilistic is irrelevant — if anything the effect of these factors on our thinking is deterministic. What matters is that the mechanical process of producing thought is clearly influenced (perhaps entirely!) by non-rational factors. To me this means that any characterization of the essence of thinking that relies too heavily on its logical structure cannot be telling the whole story.

I had high hopes for this essay because I’ve tried many times to get people online to articulate what they mean by “it doesn’t really understand, it only appears to understand” — my view is that all these arguments against the possibility of LLMs thinking apply equally well to human beings because we don’t understand the process that produces human thinking either.

But the essay is a huge letdown. The European vs. American framing obscures more than it illuminates. The two concepts of intelligence are not really analyzed at all — one could come up with interpretations under which they’re perfectly compatible with each other. The dismissal of Marx and Freud, two of the deepest thinkers in history, is embarrassing, saying a lot more about the author than about those thinkers.

(For anyone who hasn't read much Freud, here's a very short essay that may surprise you with its rigor: https://www.marxists.org/reference/subject/philosophy/works/...)

To the extent this is an accurate characterization (somewhat, I think), they considered the quality difference a bug and fixed it!

And it’s just a restatement of the original, which isn’t any good either. Really mediocre.

I tend to think it would be very hard and very risky for large, successful companies to systematically lie about these things without getting caught, and the people who would be doing the lying in this case are not professional liars, they’re engineers who generally seem trustworthy. So yes, if there is a degradation, I think bugs are much more likely than systematic lying.

To be clear, the company has very directly denied doing this.

Growing up I would see authors listing particular species of trees when describing a scene, and I’d marvel at the idea of someone getting all the references. It seemed so old-timey. But during the pandemic, my wife and I got into plants because it was an outdoor activity. I used an app to identify all the trees in our neighborhood (then we found out our town has a map online of them all). I have my favorite ones I like to go by on walks. In a given area there are really only 10 or 15 species you have to know to cover most of the trees you see. Being aware of these things adds depth to the world as you experience it.

Are we talking about the same article?

I really don’t think search captures the thing’s ability to understand complex relationships. Finding real bugs in 2000 line PRs isn’t search.

Likewise!

To me, debunking hype has always felt like arguing with an advertisement. A good read about that: https://www.liberalcurrents.com/deflating-hype-wont-save-us/

I don't get who is saying this dreaded "you'll be left behind." The only place I see that is from straight-up slop accounts in the Twitter algo feed. Surely you're not letting those people make you feel bad.

> You see a lot of accomplished, prominent developers claiming they are more productive without it.

You also see a lot of accomplished, prominent developers claiming they are more productive with it, so I don't know what this is supposed to prove. The inverse argument is just as easy to make and just as spurious.

This is the post that made jj click for me, and not coincidentally it is about a rebase operation that feels complicated in git but trivial in jj.

https://lottia.net/notes/0013-git-jujutsu-miniature.html

In about 12 years of using git (jj user now) I almost never rebased through the CLI, but I found shuffling branches around in a GUI pretty intuitive. I liked GitUp[0], which gave me undo way before jj existed.

The common view that a Git GUI is a crutch is very wrong, even pernicious. To me it is the CLI that is a disruptive mediation, whereas in a GUI you can see and manipulate the DAG directly.

Obligatory jj plug: before jj, I would have agreed with the top comment[1] that rebasing was mostly unnecessary, even though I was doing it in GitUp pretty frequently — I didn't think of it as rebasing because it was so natural. Now that I use jj I see that the cost-benefit analysis around git rebase is dominated by the fact that both rebasing and conflict resolution in git are a pain in the ass, which means the benefit has to be very high to compensate. In jj they cost much less, so the neatness benefit can be quite small and still be worth it. Add on the fact that Claude Code can handle it all for you and the cost is down to zero.

[0]: https://gitup.co/

[1]: https://news.ycombinator.com/item?id=46602056

There’s tooling like https://github.com/LucioFranco/jj-spr for managing stacks of PRs, but for stacks of 2 or 3 it’s not too bad to do it manually.

I've used both CC and OpenCode quite a bit and while I like both and especially appreciate the work around OpenTUI, experience-wise I see almost no difference between the two. Maybe it's because my computer is fast and I use Ghostty, but I don't experience any flickering in CC. Testing now, I see typing is slightly less responsive in CC (very slightly: I never noticed until I was testing it on purpose).

We will see whether OpenCode's architecture lets them move faster while working on the desktop and TUI versions in parallel, but it's so early — you can't say that vision has been borne out yet.

I was thinking of that when I wrote it.

Counterpoint: no, they're not. The test in the article is very silly.

Pretty sure I have seen them imply in one of the panel discussions on their YouTube channel (can't remember which) that they get unlimited use of the best models. I remember them talking about records for the most spent in a day or something.

I agree! I just don’t think the friction has to come from tediously trawling through a bunch of web pages that don’t contain the answer to your question.

This is just not true. I have wasted many hours looking for answers to hard-to-phrase questions and learned very little from the process. If an LLM can get me the same result in 30 seconds, it's very hard for me to see that as a bad thing. It just means I can spend more time thinking about the thing I want to be thinking about. I think to some extent people are valorizing suffering itself.

Thanks for replying! I disagree that current LLMs can't help build tooling that improves rigor and lets you manage greater complexity. However, I agree that most people are not doing this. Some threads from a colleague on this topic:

https://bsky.app/profile/sunshowers.io/post/3mbcinl4eqc2q

https://bsky.app/profile/sunshowers.io/post/3mbftmohzdc2q

https://bsky.app/profile/sunshowers.io/post/3mbflladlss26

I see this response a lot but I think it's self-contradictory. Building faster, understanding faster, refactoring faster — these do allow skilled developers to work on bigger things. When it takes you one minute instead of an hour to find the answer to a question about how something works, of course that lets you build something more complex.

Could you say more about what you think it would look like for LLMs to genuinely help us deal with complexity? I can think of some things: helping us write more and better tests, fewer bugs, helping us get to the right abstractions faster, helping us write glue code so more systems can talk to each other, helping us port things to one stack so we don't have to maintain polyglot piles of stuff (or conversely helping us not worry about picking and choosing the best stuff from every language ecosystem).

Very impressed at the rate of high-quality interesting posts from this author.

Reasonable, and I get it because I did the same thing before agents got good this year (obviously good, I say again) — I felt the trajectory was clear but didn’t want to sound like the shills and wackos.

On the other hand I think accusing Zitron of subtlety or tempering his rhetoric is a bridge too far.

I think that’s roughly right — both then and now he has stressed that people think it does something but it fails to do so. However I do think I’ve seen a subtle shift in phrasing in both him and other critics as it has become more obvious and undeniable that experienced and highly skilled experts in various domains are in fact using LLMs productively to do all those things (most notably producing software)

I dug around a bit but wasn’t able to find a slam dunk quote from a year ago. Might look around more later.

In fact I do!

"I know I sound like an asshole, but I’ve got a serious question: what can LLMs do today that they couldn’t a year ago? Agents don’t work. LLMs - read stuff, write stuff, analyze stuff, search for stuff, 'write code' and generate images and video. And in all of these cases, they get things wrong."

https://bsky.app/profile/edzitron.com/post/3ma2b2zvpvk2n

This is obviously supposed to be a critique, but a year ago he would never have admitted LLMs can do any of these things, even with errors. This seems strange but it's typical of Zitron's writing, which is often incoherent in service of sounding as negative as possible. A couple of other examples I've written about are his claims about the "cost of inference" going up and about Anthropic allegedly screwing over Cursor by raising prices on them:

https://crespo.business/posts/cost-of-inference/

https://news.ycombinator.com/item?id=45645714

This is a great read. I just want to point out what great marketing this and the WSJ story are. People reading it think they’re sticking it to Anthropic by noticing that Claude is not that good at running a business, meanwhile the unstated premise is reinforced: of course Claude is good at many other things.

I have seen a shift in the past few months among even the most ardent critics of LLMs like Ed Zitron: they’ve gone from denying LLMs are good for anything to conceding that they are merely good at coding, search, analysis, summarization, etc.