And don't forget the 'Thing' from Addams Family in the shoulder of the guy on the left.

The guy on the left in that same photo only has three fingers (and a thumb, I suppose). I thought image generation has already outlived that.

Edit: I feel stupid I didn't see the original OP already mentioning three finger issue. I'll just leave it here.

My guess would be that somewhere you have these instructions being fed to the agent. Double check skills, AGENTS.md, memory, agent definition...

You can also ask why did he mentioned something that wasn't done or why he thought this was important.

In my AGENTS.md file I have an instruction telling the agent to never commit any changes unless I explicitly ask for it, and this leads to messages similar to what you just described.

I'm starting to think that 'memory' may be the wrong analogy for what we want.

I do think that having a set of token that are highly personalized to your project and to way you work is beneficial. I also think that the idea that this set of token will be constructed in the background without any work from the user is really appealing. So it's understandable that the 'memory' analogy became so popular.

But in my experience having a really good AGENTS.md file almost always produce better results than enabling memory.

Maybe we should start to think about how we 'train'/'onboard' agents into our projects, in a similar way that we do for new co-workers. Imagine if we could send the agent to our repo and ask it to learn our patterns and in the end we could quiz the agent to gauge how much it actually understood the project. Once he 'understands' the project we can start to use it to help with development.

In a very small scale (example, individual new features) I will sometimes ask the agent to explain me how things work (even though I already know how it works) so I can 'prime' the agent context with good data before starting any real work. But I'm not sure if this approach could be reliably scaled to work with any repo for any kind of work.

When I was doing a really large refactor across the codebase I told Claude Code to explain how certain things worked currently and how I wanted things to look like after the migration and a plan on how to get there. Then I did several further clean sessions where I always started along the lines: "Using this plan {link} do X."

Works pretty well for "non-permanent" instructions (you don't want to put this info into your committed markdowns).

My main motivation was simply to save tokens, but it actually worked really well and improved speed as well.

we want some kind of fractal knowledge graph. It starts coarse at some zoom level and you can move in and out. You can insert knowledge at any level and it's out/in levels adjust accordingly. The search semantics at every level are the same, but whats in the visible area changes based on what your focus is.

One way I've toyed with a graph outline with it using "whitening" (https://arxiv.org/pdf/2104.01767v3) for embeddings rather than just text, so you add things like file path, nearest title/method/const etc. You have to have dummy text though because it fails with "null"; all embeddings need to carry some kind of text and of the same size.

so when the agent rembembers something, the memory would be an embedding that includes where they found the file, what method or const or whatever they're in, what the task they're working on is, etc. That all becomes a single embedding. You could imagine a metadata tag that also describes the tools they're using etc.

Map out a complete space of tags for whitening an embedding and there's surely a proper mix. Then when you're searching for things in the embedding, you also store some of the other metadata as plane strings & edges, which gets you some useful granularity.

It's really about storing institutional context and on-task learnings. The AGENTS.md can do the same thing as memory, but if you have a good memory system, in theory you never need to do any ongoing maintenance of AGENTS.md and the system will do it automatically. And because it's automatic and has a whole harness system built around maintaining the memories, you can engineer it to have not just more convenience of maintenance but also better correctness / recall quality.

Memory systems can also let AI load memories on demand so it's actually analogous to maintaining AGENTS.md plus twenty different "read this if you need to do X" markdown files.

I think the name "memory" makes sense given that the AI system is one singular system with central context (rather than a software org of multiple distinct humans with distinct memories), so the equivalent of institutional knowledge in the software org really is just akin to memory for the AI system.

>I suggest groupthink is a critically important aspect of most human behavior. And yet somehow, we rarely manage to discuss this, and are repeatedly surprised by it. I'm unclear on why.

Isn't that just a manifestation of the 'Blind spot' from Johari Window?

In general it is really hard to think about the things you already internalized.

Perhaps not when it's a failure to consider groupthink in others?

I don't often see something very vaguely like "Financial folks are anticipating X, but as a community they don't have a great track record of anticipating, err, the internal politics of <country>. So for that part we'll turn to ... , though that community's grasp off financial context has often been weak." etc. Systemic variations in the capabilities of communities, being systematically acknowledged and addressed, in the context of public analysis and discourse.

I really don't understand their thinking here. Sure they want more money, I get that.

But 'physical media' is one of the reasons why a lot o people make a distinction between PC and console games. Removing this will make it easier for consumers to compare a PS5 to a Steam machine, and I don't think that is a good thing for Sony.

I'm thinking about getting an Audiocast M5[0]. It seems to do exactly that. The price seems to be around US$ 30.

[0] - https://audiocast.io/

Doesn't look like the Amp50 is available to buy...

That's missing the 'amp' part.

You're 75% correct, one of the 4 products does have an amp though.

He uses the rumble to move the controller to the charging puck in a flat surface. He uses a camera to identify the direction we need to move the controller. When the controller gets close enough the charging puck automatically attaches to the controller because of the magnetic force.

Instead of training the model to directly answer questions we trained the model to always write and execute the code that would solve the question ?

If that is the case, this isn't just a fancy way to perform prompt optimization?

You could put this captcha in a location that wouldn't be very visible for a human, but if the LLM is looking at the HTML he would find this form.

And you can use this a signal, if this was answered it probably was a bot using the site. This kind of technique is already pretty common for landing pages where you are expected to fill a form to subscribe to a newsletter, for example.

Does hiding things from humans with display:none or visibility:0 work against bots. Don’t they look at the styling? Even stacked elements should be discernible.

> Computer use is such a terrible idea. It's slow, insecure, error prone, expensive.

And yet having an agent able yo use a computer on your behalf is really useful.

Recently I gave a Nix OS vm to my hermes agent and it has been a good experience. I don't really care if destroy the machine I can just rollback to an earlier version, and for any meaningful data he creates for me I make sure he creates a repo, commit and pushes to my private Gitea instance.

> And yet having an agent able yo use a computer on your behalf is really useful.

It is, but there's no need for it to be viewing your screen, browsing websites and watching ads.

That stuff is for humans, not for LLMs.

Sure, I don't want an agent watching MY screen. That's why I gave him his own environment, and pretty quickly he discovered that you can open chrome and make it render to a framebuffer, this way he is able to 'view' the website. And apparently with this he is able to bypass a lot of 'anti-bot' measures.

> And yet having an agent able yo use a computer on your behalf is really useful.

I honestly cannot think of a single use case

I think the main advantage is adaptability.

Imagine you have a pretty exotic task you need to complete that involves converting a video file from one format to another.

You can use ChatGPT or something similar and the best you will get is either a script you can run on you machine that does what you need or he may decide to render a new video.

If you have something like OpenwebUI you could configure a MCP that converts videos and allow the model to use this MCP to do your task. This should work, but is quite a lot of work for something you'll ever do once.

But if the agent has it's own environment he can decide to install ffmpg, execute the transformation and serve you the file you want.

In reality there is no new capabilities with this approach, but things get a lot more comfortable.

This doesn't require computer use, just a bash tool (and possibly fetch to get ffmpeg documentation)

Yeah even Claude Cowork would do this, doesn't need "computer use"

Literally everything you do every day.

It's the end game of AI. Have systems trained on doing EVERYTHING you do on a computer all day. Trained by you while doing the job.

Have you ever done something tedious on a computer?

I give you one: Google news is pretty terrible right now almost all interesting new sources are paywalls and so I get recommended all kind of weird lifestyle publications that are really horrible. With the computer use API I can just tell. Tell Gemini to look at Google news pick the articles that look interesting. Look them up on archive.is, and just give me the plain text article and construct a summary - I think that would probably work pretty well.