Computer use is a great idea. It gets the job done when nothing else will.

If you're a person trying to get their job done at a big company, but half your job is in 1-2 proprietary tools or is stuck behind an API you can't program against, computer use can allow you, a non-techie, to do your job more efficiently.

I think it's an awesome way to circumvent gate keepers and the IT department to let people accomplish their goals.

That is an incredibly niche use case and comes with a boatload of footguns.

Even then, an AI writing AHK scripts likely outperforms.

It does. I used to be an ahk "script kiddie" and know it front and back. It's sort of burnt into my brains. As a result, I can prompt really really well, notice issues at a glance, and I have a sheer volume of scripts locally for all sorts of tasks some from as far back as 2014. From tiling window managers to OCR all the way to simple hotkeys/hotstrings. I let it grep in that folder and build out whatever I want using those primitives. This gives actually 1-shot immediately usable 100% working scripts even with GPT3.5 level models, as opposed to the iterations needed for typical development.

Example: adding copyright text box to bottom of every slide

  F3::
    pres := ComObjActive("PowerPoint.Application").ActivePresentation
    Loop % pres.Slides.Count {
      slide := pres.Slides.Item(A_Index)
      box := slide.Shapes.AddTextbox(1, 100, 500, 500, 30)
    box.TextFrame.TextRange.Text := "Copyright 2026. All Rights Reserved."
    }

  return

Yeah, it's not that computer use is the most theoretically optimal paradigm, but there's a reasonable case that given the constraints of modern software systems and how they're built, that it's the most realistically optimal paradigm.

Computer use is very useful for developing GUI applications since claude code can build and test the entire app end-to-end (accessibility APIs exist but depending on the UI framework of your choosing you can run into walls very fast).

I run it in a VM using a headless wayland compositor, I'd never trust even fable with access to my real system.

I think there's a sweet spot- a lot of the time you're probably better off with "reverse engineer this web page and build me an API or personalized chrome extension to meet my needs".

I have an agent doing price checks for me for an item on a certain website. Instead of blasting through a zillion tokens processing the DOM over and over, it loaded the page once and figured out how to download a json with the price.

Does it have to view the page now repeatedly to download the JSON?

It curls the page. I think the approach it took won't actually wouldn't work in my local browser- its getting the value from some conversion reporting code that I'm guessing my ublock extension would hide.

The tool it built will do viewing, probably.

that tool is called Chrome with extra flags in the CLI

How are folks using “computer use” to click things on intranet portals that are behind an SSO? Even this OP example shows visitors a url and enter this search term… that is port of useless.

How can I automate things behind an SSO wall? Even if it means I manually authorize it once and watch it do things on its own..

I've never used Gemini computer use, but I assume it's the same:

Claude computer use takes control of your whole computer inputs (mouse and keyboard) plus screenshots. You just log in, tell Claude you're logged in, and let it get to work. It'll use the browser you're logged in with.

The chrome extension is a little better because it only takes control of its own chrome tabs (again: you just log in.)

Take manual control once, save the login info to a password manager, teach the model to login with it.

For anyone trying to figure out how to build a society where no one wants to be a criminal, I highly recommend When Brute Force Fails: How to Have Less Crime and Less Punishment by Mark Kleiman.

There are evidence-backed ways of reducing criminality.

One counterintuitive way of reducing crime is to increase the likelihood of being caught, to have small-but-increasing consequences for committing crimes, and to increase the swiftness of sentencing.

For example, if you are caught drinking and driving, you immediately spend 1-2 days in jail.

Long sentences are not very productive at reducing crime or at least are a very inefficient way to do so.

An intuitive way to understand it is imagining that there was a system where if you stole something, you 100% of the time got hit with a charge to your account of the item value + $10. No one would steal again even if the penalty for getting caught was relatively nothing. Because the feedback loop is so short and guaranteed.

No ones life would be ruined over a dumb choice and yet they would change their behavior very fast.

You have to deal with the judgement-proof somehow.

There’s effectively two sets of laws - one for those with something to lose (fines, etc) and one for those with nothing to lose.

It’s still the same. If you steal something and have no money you lose the item and get some small penalty, perhaps a day in jail. If there is absolutely no chance you’d get away with it, everyone quickly realises there is no point.

Scott Alexander has a good analysis of these issues:

https://www.astralcodexten.com/p/prison-and-crime-much-more-...

Curious how you tested this - did you do repeated home tests to figure this out?

Most CPAP machines including the one I have record a lot of data that can be analyzed via software like OSCAR:

https://www.sleepfiles.com/OSCAR/

On the machine I have (ResMed AirSense 10) the data gets written to an SD card I can pull out and import into OSCAR to see what the machine detected about my sleep apnea events.

I also (not always, but when I'm testing changes) use a continuous pulse oximeter to test my SpO2 blood oxygen % over the entire night. The data from this can either be combined with the data from the CPAP machine, or used on its own to test things like how I do with the collar and no CPAP machine, since the SpO2 drops are the dangerous thing I'm trying to avoid by eliminating the apnea events.

I use the Wellue SleepU for SpO2 monitoring but there are a lot of similar devices (the kind with finger tip sensors are generally more accurate than the wristwatch based solutions).

I think it does. Crops pull up a set amount of water. If it's concentrated, then they'll pull up a lot of heavy metals. If it's at very low levels, then they won't.

Is there an alternative?

We live in a much, much cleaner world than we did 50 years ago. Legislation and environmental rules have worked. There are some areas where it could obviously be better, but also some areas where regulation is too strict (blocking housing, renewables, transit) and the system is evolving to address those.

I think the loss of local media has made it harder for misdeeds to come to light, but I don't want to throw up my hands and cede everything to commercial interests et al.

I think a look at other countries would do well. There are many with much tighter regulations (e.g. EU countries, Singapore, Japan) and they seem to have good results with that.

> We live in a much, much cleaner world than we did 50 years ago. Legislation and environmental rules have worked.

I think prevention of pollution is one area where very tight regulation is absolutely needed, and this seems to be an argument for that.

Of course regulation can be weaponized and used as a tool to serve entrenched interests as well - but this is then more a problem with the overall political system. Also, I think a proof that this is the case is necessary instead of assuming it by default.

Then we should work together as a society to fix the law and make sure it's applied evenly. Hard to do, but necessary.

Is there an alternative?

This is written for the 3 models (Sonnet, Haiku, Opus 3). While some lessons will be relevant today, others will not be useful or necessary on smarter, RL’d models like Sonnet 4.5.

> Note: This tutorial uses our smallest, fastest, and cheapest model, Claude 3 Haiku. Anthropic has two other models, Claude 3 Sonnet and Claude 3 Opus, which are more intelligent than Haiku, with Opus being the most intelligent.

Yes, Chapters 3 and 6 are likely less relevant now. Any others? Specifically assuming the audience is someone writing a prompt that’ll be re-used repeatedly or needs to be optimized for accuracy.

This enables the /resume command that lets you start mid-conversation again.

Storing the data is not the same as stealing. It's helpful for many use cases.

I suppose they should have a way to delete conversations though.

That's not just them saving it locally to like `~/.claude/conversations`? Feels weird if all conversations are uploaded to the cloud + retained forever.

Ooo - good question. I'm unsure on this one.

The conversation is stored locally. If you run Claude on two computers and try /resume, it won’t find the other sessions.

They are downloaded from the cloud and just never deleted.

Resume is now a drop down menu at the top in the new VS Code plugin and it's much easier to read.

> New native VS Code extension

Looks great, but it's kind of buggy:

- I can't figure out how to toggle thinking

- Have to click in the text box to write, not just anywhere in the Claude panel

- Have to click to reject edits

I wish I could put it in the sidebar like every other flavor of AI plugin.

It seems they also removed the bypass permission setting...

plans now open in a separate file tab, and if you don’t accept it, it just…disappears so you can’t discuss it!