[alain94040]

Neither the original 1987 story nor the blog post make any sense to me. I feel the authors are focused on a niche aspect: exchanging deep human thoughts with other humans. For that use case, sure, don’t use a computer, don’t use AI.

But there’s nothing wrong with a car made by a robot factory. The car is useful, regardless of how it was made. AI lets some people create new products, those may be useful.

[olalonde]

Probably not the author's point but to me, the takeaway is that these people are driven by emotion and can't be won over with rational arguments. They have an emotional attachment to their old ways, a fear of being obsoleted, a wish that there is something truly special and mysterious about humans, etc.

[archagon]

And the other side is convinced that there’s nothing truly special and mysterious about humans, which feels inherently disturbing. Like, these people would swap their brains with a computer in a hot second if they thought it would make them do X a little better. Shudder.

[olalonde]

I don't think that's true. I think most people are naturally against this idea emotionally. Why would humans not want to be special? I know I wish it wasn't true.

[fedeb95]

the point I see is that even if you automate all that is possible to automate, there are some core activities that can't be delegated to machines in order for a product to feel very good. May it be the design, or whatever: I don't have an opinion on which aspect must be curated by a human, but some aspect has to. Maybe this aspect is simply AI review and prompting, or AI infrastructure creation and maintenance... I don't know. But the point is that this part of production exists, and it's what makes a useful product something you want to buy more than another useful product.

[bLorax]

I'm pretty undecided on where I stand on LLMs, but I have to point out that I use them before making my point. The thing with AI is that the output we get from it is not deterministic. With assembly lines you can get a good enough deterministic output (having component failures in mind) that everything is gonna be fine. I've never been to a car factory, but I'm sure there are guard rails in place to ensure minimum damage if something goes wrong. There is simply not such a thing with AI, thus the programmer has to be the guardrail. The blog post leans a bit heavy into sentimental words for the programmer's mind, but I'm guessing this is what they mean by the "deep human thoughts" part.

[Imustaskforhelp]

I do love some comments where people have nuanced thoughts on LLMs, so thank you for sharing them!

I'm more so undecided on them because of the social effects on them. I am all for innovation but social media hasn't been the ideal example either (Facebook famously has young girls who are insecure who might delete their pics and not upload them and would show them predatory ads on top of it with countless other examples)

I am one of the younger folks on this website and I honestly feel like it might've been much better for us in the past pre social-media perhaps.

Have we as a society ever truly reflected on innovation. Nuclear innovation comes to my mind in terms of arms race (ironically which I believe that Nuclear could've been really great for power production and I hope that it gets more looked into)

I am also unsure if regulation might make sense, regulation has been tried and nothing seems to work. I had a decent chat with simon once and sort of the conclusion from that and others seems to be that the cat is out of the box.

Thus we might have to think some more about the sentimental values instead as well and do proper evaluations on if it should be used and when.

The whole culture towards AI perhaps can be seen like a fake projection of confidence. Similar to how as you point out from the deterministic/undeterministic part of it from technical point of view as well. (Many) People are projecting AI to be completely deterministic when in reality it sort of isn't.