ketzo

I totally agree and think this is The New Skill of software engineering: can you steer agents well enough to get work done at the speed they will allow, while still keeping enough context/understanding to step in when it matters?

not OP but usually for me this means long verification loop; waiting 10min on CI checks, that kind of thing, rather than actual 1hr wall clock of token generation

$19B -> $30B annualized revenue in a month?

Feels like the lede is buried here!

After seeing the last few releases for GPT and Claude, I’m not sure how anyone (else) is gonna build a durable advantage on proprietary model quality.

The capabilities of the top labs’ models have improved so much in just the last few releases, and I definitely foresee a world where they gate those models away behind 1st-party harnesses/tooling.

I think the core idea here is a good one.

But in many agent-skeptical pieces, I keep seeing this specific sentiment that “agent-written code is not production-ready,” and that just feels… wrong!

It’s just completely insane to me to look at the output of Claude code or Codex with frontier models and say “no, nothing that comes out of this can go straight to prod — I need to review every line.”

Yes, there are still issues, and yes, keeping mental context of your codebase’s architecture is critical, but I’m sorry, it just feels borderline archaic to pretend we’re gonna live in a world where these agents have to have a human poring over every single line they commit.

if you're cool just writing markdown files, I really like Astro for self-hosting static content.

Meta has shown a willingness to offer 9-digit pay packages to individual researchers. Even if they completely scrap the product, an acquihire of even a handful of Manus' top engineers/scientists here is totally in line with that kind of cash.

Yes, and some people are still happier there!

Different people can have wildly different expectations for a work environment, and wildly different tolerances of social discomfort.

But pressuring the SEC to drop the civil claims does!

https://www.nytimes.com/2025/09/19/us/politics/sec-trump-cle...

> While running the exploit, CodeRabbit would still review our pull request and post a comment on the GitHub PR saying that it detected a critical security risk, yet the application would happily execute our code because it wouldn’t understand that this was actually running on their production system.

What a bizarre world we're living in, where computers can talk about how they're being hacked while it's happening.

Also, this is pretty worrisome:

> Being quick to respond and remediate, as the CodeRabbit team was, is a critical part of addressing vulnerabilities in modern, fast-moving environments. Other vendors we contacted never responded at all, and their products are still vulnerable. [emphasis mine]

Props to the CodeRabbit team, and, uh, watch yourself out there otherwise!

What does your Claude code usage look like if you’re getting limited in 30 minutes without running multiple instances? Massive codebase or something?

That is indeed exactly what the article says — I’m not certain GP is right on this.

Kenji’s original tests seem to confound this as well: every 2.5min of slicing produces steadily less juice, despite the fact that the steak’s internal temp should be rising for some of that time.

In what ways is that better for you than using eg Claude? Aren’t you then just “locked in” to having a cloud provider which offers those models cheaply?

I mean — the person you’re describing is just a ChatGPT user and essentially nothing else, though, right?

It explains OpenAI’s valuation but no one else

A really effective prompt is created by developing an accurate “mental model” of the model, understanding what tools it does and doesn’t have access to, what gives it effective direction and what leads it astray

Otherwise known as empathy

They’re probably comparing to taxis, where you can pay in cash.

Completely unsourced and the site is run by a marketing/PR/growth consultancy.

Between that and the utter lack of detail, feels like not worthy of HN front page.

“No enforcement” means people who don’t care about breaking rules will do it in brown bags

“Officially allowed and advertised” means businesses will specifically cater to people with money who will come specifically to do it

Then what explains people doing millions of web searches on perplexity/chatgpt/claude?

I’m building out a side project where I need to ingest + chunk a lot of HTML — wrote my own(terrible) hunker naively thinking that would be easy :’)

Definitely gonna give this a try!

People said much the same thing about Apple for decades, and they’re a $3T company; not a bad thing to have fans.

Plus, it’s a consumer product; it doesn’t matter if people are “presenting them as leaders”, it matters if hundreds of millions of totally average people will open their computers and use the product. OpenAI has that.

People quite like aider! I’m not as much of a fan of the CLI workflow but it’s quite comparable, I think.

OpenAI has already forecast $12B in revenue by the end of this year.

I agree that Google is well-positioned, but the mindshare/product advantage OpenAI has gives them a stupendous amount of leeway

Have you tried Cursor or Zed? I find they’re both significantly better in their “agent” modes than Windsurf.

I think it’s just fairly rare that formal verification is valuable to a business.

Obviously all software benefits from correctness — but we all know that software certainly doesn’t need to be bug-free to make money!

And if the benefits of formal verification don’t outweigh the (non-trivial!) costs — and critically, outweigh them in a business-person-legibile way — then people won’t put up the time and effort.

I agree, and one issue for developers and designers alike is that often, it’s more fun to make a UI fun than it is to make it work really really well.

But that doesn’t mean we should lose sight of fun completely, which I think in many places we have!

While this article is a little fluffy and overstated, I do share the overarching sentiment — UIs for the vast majority of products should be at least a little fun.

“Delightful” UI/UX has become a cliché at this point, but it really does make me happy to see an element of craft and intention in the software I use, and stuff like these detailed little icons accomplishes that well!

Well, what are they gonna do about it?

Nvidia has the most desirable chips in the world, and their insane prices reflect that. Every hyperscaler is already massively incentivized to build their own chips, find some way to take Nvidia down a peg in the value chain.

Everyone in the world who can is already coming for Nvidia’s turf. No reason they can’t repay the favor.

And beyond just margin-taking, Nvidia’s true moat is the CUDA ecosystem. Given that, it’s hugely beneficial to them to make it as easy as possible for every developer in the world to build stuff on top of Nvidia chips — so they never even think about looking elsewhere.

For a long time, I’ve been this person for other people, but don’t feel like I have anybody to do this for me. That’s okay — I don’t feel bitter about that or anything. And I don’t wanna overstate what a good friend I am or whatever, I just do this a decent amount. But some part of me does wish I had someone celebrating my wins.

This:

“No one comes to mind? Maybe you haven’t really trusted anyone with your wins yet.”

really, really hit me for some reason. I’m pretty averse to praise/congratulations — even if I feel it’s deserved! — so I don’t really share my wins with people. How can I expect to have people hype me up if I don’t let them in a little? It’s obvious when I write it all out but I kinda can’t believe how long I’ve been operating this way.

Anyway, great post!