cush

I never said engineers should be liable instead of the firms they work for. Both should be liable.

Your response feels like you were just looking for anything to shoehorn your overt racism into

An engineer is liable but so is the firm. A doctor is liable but so is the hospital. I’m going after both.

How would making imported workers liable do anything but disincentivize them from committing these crimes?

Their users are ad execs and data brokers

If only software engineers were liable like real engineers are

> it’s simply not humanly possible to keep up

Are you talking about manual review? Of course it isn’t possible to read that much, I never said it was. The scale of changes requires new approaches to engineering. Some may feel these processes are less rigorous, but in my opinion, when done well, they’re actually more.

I have worked in many codebases using coding agents. Sometimes their changes are large and sometimes they’re small. The better repos have many guardrails to keep the PRs small. I’m not sure why you’d assume why my or your experience working with them would be even remotely similar to working with them at Anthropic though. I think it’s just fine to say you don’t know. They have no obligation to appease to people debating in an HN thread.

I think in general you’re going to continue to struggle to translate your one-off experiences to these 1000x 10000x scales. Most people can’t fathom scaling anything beyond 10-100x.

And depending on the job, AI might replace the vital functions leaving us all holding doors

It’s kind of ironic how your comment starts with “LOC is a decent measurement” with zero explanation, then quickly devolves into rough estimates, assumptions, and finishing with “6M/day… it’s still insane”

You’re proving precisely what I’m saying. You don’t have any real information and your guess is as good as mine. It’s just as plausible to assume Anthropic runs 50x experiments for every feature they ship and only 2% of the code hits production. It’s just as plausible to assume half or more of their repo is tests. It’s just as plausible to assume there’s as much comments, skills, docs, and context engineering as there is code. It’s just as plausible to assume that there’s 10x more configs, localized strings, and content as there is executable code.

If you add all my assumptions together that leaves us with about 15x PRs worth of shipped executable code. And even with your generous 200 LoC per PR, it’s only 3000 Lines per day. Now that’s insane!

That’s fine. You can say “that’s a cool car but I’m worried it will end like the cybertruck” or “that’s a cool car and he’s a deadbeat dad”, but there are so many people in the comments saying things about how the car isn’t doing anything novel or interesting and that’s just ridiculous. It’s straight up sci-fi.

Honestly systems like this resemble real engineering more than manual human review does

30k PRs is such a meaningless figure though, you can't make anything of it. It entirely depends what's living in that repo. Adding localized strings, configs, things that change frequently, can easily move numbers like this by orders of magnitude. How big are the PRs? Maybe there's an auto-bug-fixer agent that runs around the codebase making thousands of single-line fixes per day. Who knows. The number means nothing in isolation. We used to scoff at bad managers for measuring the performance of engineers using lines of code, now we're using the same figures to judge AI. We've become the bad managers.

It helps everywhere... turns and straights - it provides downforce without wings. Wings cause drag, that push the car down but also slow it down. So the car can be going slow and still get incredible grip. On the corners the car won't slide, so it can corner at much higher speeds too. For both these reasons, this suction system was banned by most racing leagues in the 70's for being too powerful.

Does the thrust version not have suction too?

I'm no Musk fan. Like totally repulsed by the dude. But the comments here that there's nothing interesting about the demos are some serious cope. It's this kind of reality-bending bias that makes people disregard everything you have to say. When someone you hate creates something incredible, it's probably best to say nothing because any attempt to convince people otherwise will just backfire.

There is no world in which that car comes off the lot for $1M. That’s a $5M car, minimum

Even assuming it’s real and they genuinely are working on selling these to consumers (they aren’t), why on earth would you think this is built for the average person? At the absolute cheapest this car would be $5M off the lot. It’s for the ultra rich (assuming they ship, which they won’t) who want to drive it one day a year on the track

Building concept cars is a very important part of being a car manufacturer. The issue with Tesla is they take reservations as if it will ever go into production

Every car manufacturer creates concepts that they never plan on shipping. The goal is to attract engineering talent and do deep research. I’m curious at what price they’d need to sell it to break even…? Probably upwards of $50-100M

They build way more than just claude code. Their services, infrastructure, model training workloads, etc all live in the monorepo. Think of the systems required to monitor and respond to demand and scale needs. Agents are also implementing the entire SDLC (discovering bugs and opportunities, testing, maintenance), not just writing new "useless" features.

ScionHub?

They have one company-wide monorepo

I wouldn’t assume this is the only thing they’re doing differently

At 30000 commits per day this sounds like an integration nightmare

I think about this any time a software engineer thinks they're helping Deaf people when they build some AI thing to see/speak finger spelling thinking they're interpreting ASL. Like just ask your customers what they want

Imagine being in the thick of a global data center controversy and thinking this is a good idea

It doesn’t. Same there

Right now, LLM rewrites like this work if and only if there’s a very robust suite of tests to code against. Otherwise the port is uncanny. Enterprises are willing there’s just often too much complexity and not enough verifiability to make it happen.

Okay but we know that high resolution images, a few buttons, and complex layout are enough to bring a relatively simple web app to <60 fps. There's a legitimate story here about how we can start running desktop applications on the web. We're getting close to the point where our desktop apps can cross-compile to web rather than the other way around and I'm here for it.

I think this is terrible advice. Going into computer science right now is super risky. If anything maybe learn real Engineering, not Computer Science

It's so smooth

They're suspending the PERM process for H1-Bs but are they suspending microsoft from filing for new H-1Bs? It's my understanding the wait list for Indian workers to get a green card is like 100 years, so what difference would it make anyways?