drusepth

I don't know when you last tried agy, but if you ever go back to try it again, you'll hopefully be happy to know it does indeed support skills, sub-agents with pretty good inter-agent communication, and probably more.

edit: removed persistent memory from list since I realized I'm using a plugin for that and it's apparently not native

What basic features are missing from agy? I've been using it and cli-cc + web-cc for months (among a few other random harnesses to test here and there) and they all seem roughly comparable to me.

I actually just cancelled Ultra also because I couldn't subscribe to a YouTube Family plan while I had it active (Google... :[) but trying to use Codex as a replacement while I testdrive Astra makes me yearn for agy again.

It's almost certainly the software side. Expect "dot" pendants or "dot" pocketwatches in a year or two.

Indeed, I've been coding for ~25 years (competitively and in open source for many of them) and I max out at least two subscriptions' worth of tokens every week across a half dozen active projects.

I still code "by hand" sometimes (mostly Ruby/Rails, C#, and random languages for code golf) but just for fun at this point. Serious projects started being 95-100% AI over a year ago.

But only if you promise to spend the billion dollars on something NVIDIA is selling

I'm always curious why paywalled links aren't outright banned on HN. How is the average person supposed to read the article before commenting?

Which humans were arrested or held accountable for the Minab school attack?

True, but it's also not your job to understand what post-docs are suggesting (hopefully), unlike post-docs who kind of need to understand what LLMs are suggesting.

I think this is a severely outdated take that was true a few years ago, but has not been true for at least 1-2 years. Especially from the top models at the top labs.

Qualitative measurement is always hard, but plenty of studies have come out in the past year that generally find many models' code to be roughly equal to human code in terms of maintainability, complexity, readability, etc.

I think it's most likely the big labs have something close enough to what they'd call[0] "AGI" that they're now calling for regulation to lock up the market and prevent competition.

[0] Whether they've actually achieved some definition of AGI is way less important in this context than whether they can /sell/ what they've built as AGI.

> with vibe coding you can't cast aside the AI after you've taken the initial step

Sure you can, why wouldn't you? Good models these days spit out architecture and code fairly indistinguishable from your average developer.

It's always hard to jump into an existing project and understand it all, but it's doable and in 2026 it's no harder to jump into a vibed project than it is to jump into any other project (with exceptions in both directions, of course).

You still have the code and can switch back to "fully human" (or any other point on the spectrum of human-ai) any time you want by just opening the same editor you would use if you cloned any other project.

> I would argue that if you belong to group one you have not actually "made" anything.

That's exactly the sentiment I'm referring to.

I'm excited that group one is full of people who couldn't previously "make" anything with code, but now can (in the same way I'm excited that more people can now "make" music, art, stories, movies, etc). And they're definitely making stuff (friends with kids are making games that their friends play, sharing homemade videos with each other, making bedtime stories, making websites and apps that others use, etc).

I'd also argue that theory and architecture have always been closer to "software engineering" than coding ever was, but I also acknowledge I've got a bit of a bias belonging to group 2 (otherwise: been coding for ~25 years).

I don't know about online dating, but I think the vibe coders I've interacted with the past few years break down into two distinct groups:

1. People who let agents drive designs/plans and just let them rip (what I would /actually/ call vibe coding) -- who place value in the final product, and aren't shameful about the tools used to get there as long as the product is "good" (and generally reroll entire apps from a single prompt until they get something that's actually good)

2. People who provide (or collaborate with AI on) designs/plans and use AI to write the code (which I wouldn't call vibe coding, but often gets conflated) -- who place value in the planning, and aren't shameful about the tools used to implement them as long as the architecture is "good"

Either way: I'd say most people who "vibe code" (in either category) are a little annoyed at the term inherently belittling what they make, but generally dismiss the term rather than feel any shame about it.

I'd be surprised to learn anyone in my programming network doesn't use any AI anywhere. It's a very powerful tool that's integrated pretty much everywhere into the 2026 Developer's toolkit, and the question is generally /where/ you use it (or not), not whether you use it at all.

I mean, that post /is/ entirely about AI.

The "I love AI" posts are just as bad as the "I hate AI" posts, and the same "this is my opinion on AI" posts are as bad as the onslaught of poorly-AI-vibed demo posts. Makes sense to filter them all out IMO.

Same here. I assume this doesn't work in Chrome or something?

Yes, but why?

Yes, but presumably they'll work on another problem instead, because they're mathematicians who enjoy doing mathematics.

Is there value lost in them working on problems that don't have solutions instead of problems that do?

I have a hunch the real killer feature of assistants like this will actually be the ability to identify moments like this where it'd actually be helpful, instead of relying on users to actually try/do everything with it.

Ironically, I feel like LLMs have really unlocked Windows work since they all know how to use Powershell and all the weird Windows commands/hacks that you otherwise either wouldn't even know about, or would have to Google every time.

I used to hate working in Windows, but now it feels roughly comparable to Mac (though at the risk of inciting an OS war, I do feel like Linux still beats both out...)

> GLP-1 receptor agonist medications typically cost between $149 and $350 per month for cash-pay oral pills, and $900 to $1,400+ per month for list-price injectables without insurance.

LLMs started meaningfully passing the Turing test a year or two ago, around GPT-4.5. Is there another version or bar for "passing" you're looking for?

[0] https://arxiv.org/pdf/2503.23674

> I asked Claude to fix the grammar of my comment

I read this to mean he wrote the comment, then asked Claude to fix the grammar (as many ESL speakers do). Sounds to me like he did write it.

I've curated my auto-approve list to specific commands by approving them with "always allow in this project" (I never want it e.g. committing/pushing to GitHub, removing files, etc without being in the loop) but Antigravity does have both a standard "auto-approve" mode _and_ a "Turbo mode" which disables ALL approvals of all kinds.

Is there a reason these pelicans always have roughly the same composition (side-view, 2d, biking right, flat ground beneath, etc)? I don't see any of that detailed in the prompt, yet they all seem to generate roughly the same image of differing quality.

Antigravity is probably the best of the bunch I've tried. I'd say it's pretty comparable to Claude Code (I use both daily).

Interesting side note: although Opus is still image-only, you can still drag videos into Claude Code and it doesn't blink an eye; it just strips it down to a series of images to parse.

True multimodal support would be way better, but I have no issues pasting in full screen recordings while QA'ing games and having Claude identify and fix issues in the video.

Once you've written something, it's incredibly easy to overlook minute changes to the text.

See: why authors wait days, weeks, or even months before editing what they've written (or, if you're more interested: cognitive regression, inattentional blindness, and the effects of misdirected saccades).

What browser are you using that doesn't have feet (or animations)? Seems to all be there in Chrome.

What should you be looking out for as bad behavior from this model? I've daily-driven 3.5, 3.6, and now 3.7 for however long it's been out (alongside Opus, as I hit daily limits on both subs each day) and haven't noticed anything "untrustworthy".

A huge benefit of local setups at our studio is that a lot of our software can't run headless, so when we're having agents work in Blender or in Unity etc with MCP that'd otherwise eat up our normal computer use. (Try to have two people try to work in the same Unity editor at once... then try ten!)

We also built some QA agents that are always playing our games from the same builds a player would and flagging things to fix/improve; that alone needs the game focused and front-and-center so it can properly screen-capture for deciding what inputs to take next (and for screenshots/replays), which also means we can't really do any hands-on work at all on the machine when it's running.

Having a separate (and tiny) machine for all of this has been great. We don't bother with local models because, you're right, the $20/month sub is way better than anything that can run on small consumer hardware atm.