AI enables curious people to explore. Why do you say it kills curiosity? If anything, it's so recognizable with output I'd say it kills creativity.
Whenever I tried the same with Google in the past, more often than not I couldn't find what I was looking for, because I didn't know the correct keywords to search for in order to start getting relevant results. With ChatGPT & co. I can just pose the question in natural language, get results and continue exploring.
So I guess I'd say it's more about how you're using the tool and what kinds of problems you're looking to solve with it. A calculator can be dinged for getting effortless answers at every turn or it can be praised for enabling a higher volume of solved math problems and enabling more complex work for a broader set of people.
If you're just copy-pasting answers and you don't internalize what is being said, sure, you're not being curious or more importantly, learning. This DOES NOT mean that every person who engages with an LLM is doing that or doing it every time, and just like using a search engine or grabbing a book can lead you into interesting rabbit holes, so can an LLM, it's just a matter of how fast and to want end.
The real issue is the hallucinations which for people unfamiliar with said topic, can lead them into believing what they're being told is a fact when it's not. Also LLMs like leaving out URLs and sources from their replies to save on tokens often if you don't remind them, that's also annoying.
This whole discussion is bunch of anecdotal evidence, which is fair, and as such I'll give my own. I've found myself engaging more with obscure topics that interest me via the LLMs than I did with a search engine because the barrier is lower. I don't have to sieve through horribly designed websites filled fluff that doesn't interest me, many with dozens of JS trying to run (UBO + noscript thumbs up) and in some cases demanding that certain JS run just for me to see some plain text, some slow to browse with topics hidden under sub-sub-menus. It's annoying and just one of many barriers. Others being language. etc...
The real problem is that most people either don't see the value in or don't have the time to indulge in their curiosity. Even the language we use, indulgence to describe scratching that itch. How funny. Because curiosity is a luxury.
It is indeed. Curiosity, for me, very often stems out of a particular kind of idleness and boredom, paired with a tricky question I can't find an immediate answer to.
And I can definitely still be bored that way even with LLMs.
Do you never click through to the sources or experimentally test the information presented to you by the LLM? If not, who's stopping you? To me, this seems a bit like a tenured academic complaining about the abundance of research assistants working for them preventing them from properly understanding things anymore.
You think things like "is the accordion a better user experience than the side tabs" instead of "why the f is the third accordion pane empty?"
Sure, the curiosity of figuring out where you made the mistake is gone, but that was never very valuable. It's just a detour that forces you to be curious about something else.
Fun times.
I brought it up two years ago and get downvoted when I brought it up a couple months ago.
There is a story on the front page right now about someone losing their child's family videos from a youtube ban. We hear about this stuff all the time. I suspect we are gonna be in somewhat of an arms race with AI products as the bubble grows over the next 18-24 months. This makes me worried about how disadvantaged people are going to be if they lose access to the better platform (whichever that ends up being).
Do you think AI is going to be so important that we would benefit from legal protections for access?
Or do you think the models and technology will become so small we will be able to personalize / decentralize the tech and it still be useful / competitive?
These things can be a poisoned chalice, leading to weaker long-term performance, or they can be a force multiplier. It's up to you how you use them.
The RL metaoptimisation clearly sometimes pushes it to "here's one solution, end of story".