Yeah, no way that is a serious statement. My Macbook Air from 2011 still works perfectly, and the original iPhone that I found stashed inside one of my dad's cabinets charged, turned on without issues and was fully usable as well. That's a device that's almost two decades old. If that's not hardware quality, I don't know what is.

Having tried local agents just two weeks ago, the parent poster is correct: they don't come anywhere near frontier models, despite what the benchmarks state. I haven't tried Qwen 3.6 yet, but the version before it frequently got stuck even on moderately complex problems.

Same experience for me. I think people need to start providing context for the type of work they're doing when repeating the local model hype. Maybe they're working with a cookie cutter React app and it does the job fine.

I work on Go and Rust mostly. The experience can be wildly different based on model, quantization, and harness. The fact that it didn't work for you doesn't mean everyone is trying to hype local models, people are getting real work done.

Even when it does show the PR list, it doest necessarily show all the PRs in the category being viewed. Truly nasty issue.

Aside from a false start with Apple Intelligence, Apple did not try to repeatedly and shamelessly shove AI down everyone's throats in all their products and services, which is why their "growth" hasn't been as pronounced as those of the others. And, frankly, I'm OK with that.

They do like to pretend Siri is comparable, though. Which is, frankly, insulting.

>> You let them write code that runs in prod, which is the same thing with extra steps.

The “with extra steps” is doing a lot of work in that sentence.

The HN title is heavily editorialized. Actual article title is far less controversial: "A.I. Should Elevate Your Thinking, Not Replace It"

Ah, I was thinking of the editorialized HN title.

>> It's much like climate science today: any dissent at all, even just questioning the predictions of catastrophe, immediately brands you as a heretic.

Nonsense. It is actually quite unlike climate science, where the consensus of catastrophe and the evidence for it are both overwhelming. Dissenters are listened to only to the extent they can provide overwhelming evidence to the contrary, which they so far cannot.

>> followed by their Mythos stunt

"Stunt", eh?

I use both Cursor and Claude Code, and yes, the latter is noticeably slower with the same model at the same settings.

However, it's hard to justify Cursor's cost. My bill was $1,500/mo at one point, which is what encouraged me to give CC a try.

All corruption is bad. Selective enforcement of the law is worse. It increases corruption by giving a strong incentive to win favors from powerful people.

Do you think this guy's chances of getting away with it would have increased if he solicited favors from powerful people?

A pardon costs ~1M, just need to still more than that and you're golden.

At least they're still pretending to not be corrupt.

Inequality codified into the law, literal separate rules, is worse still.