> I'm incredibly disappointed when I sat down to my hobbyist programming time and realized copilot was suddenly and dramatically changed in a way that is incredibly disheartening.

Guess it’s time to rediscover the lost art of programming without an LLM.

I think there’s a lot of truth to “the AI did it” though. We’re encouraging the same people who get tricked by “attached is your invoice” emails to run agent harnesses that have control of your desktop. I think there’s gonna be a lot of AI-powered exploits in the future.

> I prefer humans can make choices for themselves and have reasonable autonomy when it does not have a negative affect on others.

How far do you want to take this? Your choice of diet may have a negative effect on others by way of having to pay for additional medical care.

Is taking concepts to logical extremes a good way to govern?

(No.)

But are you saying we don't care if things have negative effect on people? If we go to extremes, well then obviously everyone should have 100% autonomy? Oops that doesn't work.

So, this is the hard part - you have to find balance, compromise, a reasonable middle ground. That's always going to be the hard part. Not black or white, but the grey areas.

You're not going far enough. We need to mandate exercise.

“It’s been a whole year or two and nothing bad has happened, checkmate doomers!”

It’s pretty shocking how much web content and forum posts are either partially or completely LLM-generated these days. I’m pretty sure feeding this stuff back into models is widely understood to not be a good thing.

> Yes but that’s why you ask it to teach you what it just did.

Are you really going to do that though? The whole point of using AI for coding is to crank shit out as fast as possible. If you’re gonna stop and try to “learn” everything, why not take that approach to begin with? You’re fooling yourself if you think “ok, give me the answer first then teach me” is the same as learning and being able to figure out the answer yourself.

> Benchmarks suggests they are comparable

The problem here is people think AI benchmarks are analogous to say, CPU performance benchmarks. They're not:

* You can't control all the variables, only one (the prompt).

* The outputs, BY DESIGN, can fluctuate wildly for no apparent reason (i.e., first run, utter failure, second run, success).

* The biggest point, once a benchmark is known, future iterations of the model will be trained on it.

Trying to objectively measure model performance is a fool's errand.

> There's never been a better time to be in the trades.

I imagine it was _slightly_ better when you didn't have to worry about unlimited numbers of illegal immigrants ready and willing to undercut you.

> This is about the $10/hour that they give to freelancers to solve math/physics/chemistry problems so that they can train the LLMs on them.

If you're talking about DataAnnotation, specialized stuff like that is $30/hour and up.

It seems pretty obvious that GP wasn’t saying $10/hr to mean literally $10/hr, but were exaggerating to imply that people were getting chump change for this work. $30/hr is still chump change and not enough to buy oneself any reasonable quality of life for the majority of the population.

Well, it's something. PhD students have always been pretty poorly paid because there's a massive oversupply of them. At least they have an additional source of income available to them.

And in a public transport-dominant society every neighbor is also someone in your way, taking up a spot at the restaurant you walk to, filling up the subway train and therefore making you late in your commute…

There’s no free lunch. Doesn’t matter where you are, more people = more crowds.

Humans are pack animals, not flock or herd. I think going beyond the Dunbar number is possibly the thing making people grumpy in highly dense areas. If people cannot know those around them in a meaningful way, do they even view them as human?

What kind of meaningful connections can you make when you live in a city with a a large transient population and are surrounded by hundreds or thousands of people on a daily basis, come on now. People can only have so many friends and acquaintances. Just because my neighbors live across the street doesn’t mean it’s impossible for me to talk to them. I have about a half a dozen people in the neighborhood I know well and have work acquaintances and long term friends. Which is plenty. This whole suburbia is isolating thing is being a bit dramatized here, sheesh.

I think you have that backwards. Agentic coding is way more demanding than simple chat. The request/response loops (tool calling) are much tighter and more numerous, and the context is waaaaay bigger in general.

In processing power, but chat is interactive. Agentic coding, you come up with a plan and sign off on it, and then just let it go for a while. It's the difference between speed and latency.