What did we do before "platforms"? This wasn't that long ago (20 years give or take). If you want to share, join a club, meet people, do your hobby with a group. Personally I create for the joy of it so I don't really care what's going on on "the platforms".

> The jury is still out on whether the advantages and opportunities of AI outweight all the negative sides.

The jury is extremely clear on the benefits of AI, even if it is very annoying and being pushed everywhere. We are not going back to a world without AI, and this is something you just have to accept.

> Why does it have to keep that way?

Because you can't put the genie back in the bottle.

> Why should artists have to put up with this gross violation of copyright on a massive scale?

I don't think this is necessarily true anymore. Adobe is training their models purely on legally licensed material, for example. Many of the open source models aren't available for commercial work by default. My guess is the copyright issue is not why people hate AI. People hate AI because it's a replacement for humans and for human creativity, which sucks. But it can be legal.

> Did anyone ask us if we actually want this?

Who is us? I want AI models. Lots of people want AI models. It's not all nefarious. You can have some fun with AI as a hobby and see what it enables you to create.

> As the people we always have the right to say: "this is not ok and we demand change"

Okay, what kind of change do you want?

I totally disagree with the comparison to something like NFT. While AI is being pushed aggressively and it can definitely be annoying, AI is actually useful unlike NFTs.

Much like the author, I also enjoy photography, graphic design, and other creative hobbies. It's entirely my choice how much and where to apply AI.

We have to accept that yes, it's useful, and yes you can definitely produce good deliverables with it, not just slop. Yes, when looking for assets and not the artistic process, many people will use AI and the cost of creative work will plummet. It's not great but that's the way it is.

But it's not like we should just stop creating, especially if it's a hobby. Do it for fun. Or, maybe use AI to try something new.

Either way AI is too useful, it's here to stay and it doesn't need a do-over. It's true, we should accept the world is changing, and no amount of moaning or complaining will make this disappear.

Not everything in Israel is about or related to Palestinians. The Palestinian bias only exists in circles where every thought regarding Israel is immediately evoking a Palestinian connotation. In reality, most Israelis never interact with Palestinians.

To suggest that a sector of Israeli startups exists on the experience of people "suppressing Palestinians" is definitely biased, absurd, and is a slippery slope.

Even if you are already wealthy and don't actually need to work anymore, going off the grid completely is still the wrong move. There's a lot of ways to spend less time online, improve your privacy and reduce tracking, and still benefit from some of the actual, real advantages of tech.

And the last and maybe most important thing is, we are currently on a roller coaster of disruption and frankly some daunting prospects - but we don't know what's right around the next turn. What the development landscape might be like in a few years, or maybe what kind of new problems will emerge that are not yet clear.

The right move is to take some time off, clear your head and decide if you stopped liking tech altogether, or you just needed a break. If you still like problem solving, limit your AI use, stay effective and skillful, and find ways to enjoy your skill.

I've never met an engineer who actually stopped enjoying problem solving.

To be honest, I did have some very in depth conversations with religious friends, and it was enlightening. I am convinced that for them, it has some benefits.

But I know all this stuff is totally made up, so even if you can channel it to a good thing, I personally can never do it. And even though I won't say this to their face, part of me thinks, how can you believe this stuff.

By then we might not even have computers anymore, or we might have "transparent" computers, i.e. have everything on the cloud and just tell our AI agents what to do.

Sorry Pope Leo, things are not going to suddenly turn into a wonderful utopia, but maybe buy some stocks so you can at least make a buck from what's coming.

Also, you could spin up your own educational agent with very strict instructions on guiding the user instead of just doing the work. Of course you can always go around it but if you're making an effort to learn, this is a good middle ground.

With that said, this caught my eye:

> AI gravitates toward single-struct-holds-everything because it satisfies the immediate prompt with minimal ceremony.

This is too general. "AI" is used here as a catch-all, but in fact, it was the specific model under the specific conditions you ran your prompt, including harness, markdowns, PRDs, etc. So it's not fair to say "AI does X!" in this case.

It's also very much up to you. It's very common to have a frontier model plan an architecture before you have another model implement code. If you're just one-shotting an LLM to do everything you get mediocre, more brittle code.

This stuff is still being figured out by a lot of people. But I feel the core of the issue is not using AI well. Scoping, task alignment, validation, are crucial.

The only solution is in person meetups, bringing back the 3rd places, joining a club. Maybe it's not such a bad outcome.

> I’ve thought about this a lot over the last few years, and I think the best response is to stop.

This is exactly where it shows.

LLMs, agents and whatever comes next are not only the future of tech, but they are going to be national resilience drivers for the countries that will be able to support them with power, water and science.

Who is supposed to stop? The US? China? Russia? Everyone? Of course this won't happen. This is an arms race.

But even if it weren't, stopping is the wrong answer. You don't have to outsource your thinking, writing or reading. How you use LLMs is entirely up to you.

There is a way to use LLMs which is beneficial. I treat them as a private tutor available to me for questions. This solved a lot of friction I had with my relationship with LLMs.

More telling is that the author mainly thinks about their relationship with LLMs while in reality the space has moved on to automation with agents. You don't interact with LLMs as much as before, and if you still do, then soon you won't.

Ahents are not really ML. It's harnesses and parsing and memory and metrics. It's software. Should we stop this as well?