I can recall reading human-authored text like this for more than a decade.

But literally any decent agent can recommend existing services and help you set them up. And even help you help them set the services up for you. I do this with Claude all the time.

IMO we should all be asking for a raise if our company is making more money. Proportionally, even.

> We don't sit around and write specs and then hope working code plops out.

So what do you do then? Sit around hand-holding an AI agent while it implements code line-by-line? I'm being facetious, but my point is that if you're not doing some form of spec-driven development (that is, writing a plan and then letting an AI agent implement it mostly autonomously), then you might be operating at a slower pace than you think you are.

What primary sources are you referring to? Come with receipts next time instead of just vitriol.

None in particular. I was talking about "primary sources" as a category. See also what I wrote here:

https://news.ycombinator.com/item?id=47178573

It was a meta point. Sorry if I gave you the impression that I was weighing in on the particulars of jmyeet's essay. Rather, it was a high-level point that if you know a ton of little facts but you're only seeing half of the story, then you need to improve and broaden out your intake.

I would have the same opinion of a poster who was so one-sidedly pro-America and anti-China.

It doesn't have to reduce understanding. It completely depends on how you use it. See, for example, this study

https://arxiv.org/html/2601.20245v2

Published by Anthropic. It's a bit like a "study" by Coca Cola "proving" that one can lose weight by just drink their product rather than doing sport. Sure, it's not impossible, but that's definitely not the normal usage.

I agree some skepticism is warranted. However I think we need to avoid essentialist thinking about what effects AI usage will have on a person.

> that's definitely not the normal usage

The way I look at it, AI use -- proper AI use -- is something that needs to be taught / learned. It's not unlike other "computer literacy" skills. There are ways of using it more or less effectively to achieve your goals.

Which is so much better because you can do other terminal stuff and you can avoid vendor lock in.

That's not what vendor lock in means. If you sign up for a cloud hoster and then build your whole product on propriety services that you can't get anywhere else instead of using an off the shelf database or open source software, that's vendor lock in.

If you'd have to switch to a different tool to do your coding that's not vendor lock in.

In this case you are locking your workflow to the vendor's solution.

Did you tell it to consider future usage? Have you tried using it to find and remove dead code? In my experience you can get very good code if you just do a few passes of AI adversarial reviews and revisions.

Are you being sarcastic?

Like sure, it is. But when I'm out doing something and she texts me a book title and author, I'd have to make a mental note to take care of it next time I'm free. It also means having a stack of epub files in my phone/tablet/laptop downloads that I've got no use for.

Not only that, but the average reader will interpret the title to reflect AI agents' real-world performance. This is a benchmark... with 40 scenarios. I don't say this to diminish the value of the research paper or the efforts of its authors. But in titling it the way they did, OP has cast it with the laziest, most hyperbolic interpretation.