zingar

This is a refreshing take but I’d really have liked an example for contrast.

I find misapplication of anti-smell techniques a pretty cheap indicator that I’m looking at LLM garbage. I think they’re not really usefully engaging with that stuff yet.

How do you have the “modify LLM state from within” working? I can have it modify my config but I don’t know how to get it to eval and improve arbitrary elisp.

Absolutely baffled too. I was expecting that they preferred the vim philosophy of small tools that do one thing well, but no. So you like modal editing, well you’ve got it right there in emacs. Why that of all the potential gripes you might have with emacs?

I’d like a concrete example on how you’re actually controlling emacs with LLMs. Is ECA the part that does that?

I think there’s a weaker claim that holds true: we were able to ignore lots of content based on the superficial (and pay proper attention to work that passed this test) and now we are overwhelmed because everything meets the superficial criteria and we can’t pay proper attention to all of it.

Chess players learned to exploit chess computers’ weaknesses in the beginning too, but they can’t any longer. This version of the robot might not learn continuously, but the next will be better.

We have confidence in the extra code a compiler generates because it’s deterministic. We don’t have that in LLMs, neither those that wrote nor read the code.

What's your hypothesis about the relationship between TODOs and action?

How the neural networks produce such surprisingly human characteristics is an open question with a ton of research going into it. Explaining this is a bit more than what one smart person can achieve.

I have the same experience of reversing intentional steps I've made, but with Claude Code. I find that committing a change that I want to version control seems to stop that behaviour.

Long context as disadvantage is pretty well discussed, and agent-native compaction has been inferior to having it intentionally build the documentation that I want it to use. So far this has been my LLM-coding superpower. There are also a few products whose entire purpose is to provide structure that overcomes compaction shortcomings.

When Geoff Huntley said that Claude Code's "Ralph loop" didn't meet his standards ("this aint it") the major bone of contention as far as I can see was that it ran subagents in a loop inside Claude Code with native compaction; as opposed to completely empty context.

I do see hints that improving compaction is a major area of work for agent-makers. I'm not certain where my advantage goes at that point.

Interesting that what you're talking about as ASI is "as capable of handling explicit requirements as a human, but faster". Which _is_ better than a human, so fair play, but it's striking that this requirement is less about creativity than we would have thought.

The work where I've done well in my life (smashing deadlines, rescuing projects) has so often come because I've been willing to push back on - even explicitly stated - requirements. When clients have tried to replace me with a cheaper alternative (and failed) the main difference I notice is that the cheaper person is used to being told exactly what to do.

Maybe this is more anthropomorphising but I think this pushing back is exactly the result that the LLMs are giving; but we're expecting a bit too much of them in terms of follow-up like: "ok I double checked and I really am being paid to do things the hard way".

Fascinating. This is invisible to me, what anthropomorphising did you notice that stood out?

The article makes it seem like the author expected this without emptying context in between, which does not yet exist (actually I'm behind on playing with Opus 4.7, the Anthropic claim seems to be that longer sessions are ok now - would be interested to hear results from anyone who has).

> Maybe we should just commit the signature change with a TODO

I'm fascinated that so many folks report this, I've literally never seen it in daily CC use. I can only guess that my habitually starting a new session and getting it to plan-document before action ("make a file listing all call sites"; "look at refactoring.md and implement") makes it clear when it's time for exploration vs when it's time for action (i.e. when exploring and not acting would be failing).

I think the author is looking for something that doesn't exist (yet?). I don't think there's an agent in existence that can handle a list of 128 tasks exactly specified in one session. You need multiple sessions with clear context to get exact results. Ralph loops, Gastown, taskmaster etc are built for this, and they almost entirely exist to correct drift like this over a longer term. The agent-makers and models are slowly catching up to these tricks (or the shortcomings they exist to solve); some of what used to be standard practice in Ralph loops seems irrelevant now... and certainly the marketing for Opus 4.7 is "don't tell it what to do in detail, rather give it something broad".

In fairness to coding agents, most of coding is not exactly specified like this, and the right answer is very frequently to find the easiest path that the person asking might not have thought about; sometimes even in direct contradiction of specific points listed. Human requirements are usually much more fuzzy. It's unusual that the person asking would have such a clear/definite requirement that they've thought about very clearly.

Also the exact model/version if you haven't already.

"It's more powerful and easier" is a great claim, but I need examples in this opening page to convince me of the pain I could save myself or the awesome things I'm living without.

Any chance of a blog post covering what you saw?

Hard perhaps but it feels a lot easier now than three years ago. Or so my backlog of personal projects outside of my most familiar stack would suggest.

Parent is not making a claim about benevolence, merely about a soft power incumbent that is about to be replaced.

And (d) not worry about toddler agents wrecking your single point of failure beefy desktop

Optimising to keep the coding going 24/7 feels like a local optimisation trap. The amount of code that can be written by coding agents in normal working hours dwarfs what humans can productively describe and assess.

My efforts will be in improving agentic requirements gathering and assessment.

Who are the early access users who were providing the problems that are fairly likely to have elicited concerning behaviour?

(Apologies if this is in the article, I can’t see it)

How is evil mode support?

Hmmmm… seeing this workflow makes me wonder if I can do this with Ruby (the integration between agent and repl)

You’ve correctly identified that naming isn’t sufficient for all communication. Name the things that stay constant in the code and explain the things that vary with a particular implementation in version control messages. Version control as a medium communicates what context the message was written for, which is far more appropriate than comments.

The claim is astonishing, given emacs’ continuous use and open source scrutiny for decades. Edit: and turns out to be a problem with git, not emacs.

OTOH it’s really just the core that has been used so widely and so continuously for so long. This integration with git will have been scrutinized far less.

As an emacs user I frequently find myself in territory where I’m seemingly the only person in the world with my use case. In fact that’s half of the value: I can make emacs do whatever I want. Which means there’s security consistent with a bus factor of 1.

I don’t understand the connection to the post, could you elaborate?