If I have a robot that talks like a human, including answering questions like a human, behaves like a human in any way that is important... wouldn't this "tool" be intelligent, just like a human?
How are robots different than ChatGPT except having a physical presence and being able to manipulate the world physically?
Or are humans so special to you (or intelligence so subjective) that you can't possibly answer "yes" to the above statement about a robot being intelligent?
The analogy is more like: if robots can write like a human, then robots have the ability to write, and saying otherwise is just applying a double standard.
Similarly, if a robot can behave as intelligently as a human, then such a robot is intelligent and I don't see how anyone can argue otherwise without logical fallacies.
but are we saying robots can behave intelligently as a human?
I believe that ChatGPT can behave more intelligently than most humans at many important and useful tasks already.
I also believe that it behaves less intelligently than most humans at many tasks that are also important and useful.
For some other tasks, it behaves at more or less the same level as a human.
Moreover, I believe that the gaps between what models like ChatGPT can do and what a human can do are closing extremely rapidly.
To wit: I asked ChatGPT to write me a function to generate Hanukkah dates. It spewed a couple of facts about Hanukkah, and the generated function just returned the 8 days following Christmas. I said it was wrong, and it said "you're right!" and then gave me another wrong function (always returning March 25th, since it's actually in the 3rd month of the Jewish calendar, duh). I told it that it was wrong again and asked specifically for a Jewish to Gregorian calendar conversion function, and it gave me some more code that I have no idea is correct or not. So I asked for a table of 10 test cases of Jewish dates and their corresponding Gregorian dates, and not a single one was valid.
I actually have worked with some remote contractors who pull this kind of crap, just throwing garbage over the wall repeatedly until I give up because I just don't want to deal with them anymore. And no, they never get any better. They are not junior devs you can mentor or train or whatever. Their whole model of work is to generate something complicated enough to seem plausible and then bullshit you into accepting it so they can get paid. I have to inspect every line of code they write and do all the diligence myself anyway, and this is more time-consuming and less fun than just doing it myself in the first place. At best it's a nerd-snipe motivation in a "wrong answer on the internet" kind of way.
Strong No Hire. And if you hire them and stick me on code-review duty, I quit.
Are you saying that for current AI, or also for future AI?
I'm not an expert, but I think reinforcement learning is a pretty well-established AI technique at this point, which does give an AI incentive to be correct and become more correct.
I've heard rumors OpenAI is even doing that with ChatGPT, based on human feedback (with the thumbs up/down button).
> Strong No Hire. And if you hire them and stick me on code-review duty, I quit.
I mean, I agree. I wouldn't hire ChatGPT to replace a programmer currently, just like I wouldn't hire many humans, even many who claim they are programmers/engineers.
But at this rate, I might just hire a future version of it :)
And not only that, what's most amazing is that it can exceed humans at abilities that many years ago, most people thought only humans could accomplish (creativity, creating art, writing poems/literature, etc). And these models can only improve in the future!