0
$\begingroup$

There is no known way of building intelligence/superintelligence today (often and hereafter referred to as AGI), nor is there a known way towards building it in the future. Unlike fusion reactors, for example.

Nonetheless, certain actors have been highly effective in convincing the public that society is about to be taken over by AGI. People are increasingly afraid making objectively poor life-decisions.

I assume one of the roles of the scientific community in society is to also safeguard the masses against (religious/pseudo-religious) eschatological entities. Therefore, are there any proven or stipulated reasons for why it might be in the interest of academia, through action or omission, not to do so?

While it can be enjoyable to creatively speculate from time to time, please try to stick to the facts and try to back up your answer with sources. Also, while that may be difficult for some, please don’t share your opinions about whether AGI might or might not be possible.

Thank you!

$\endgroup$
1
  • $\begingroup$ I think this is just the bias towards money and loudest speaker, because there are many AI scientists claiming exactly what you ask, that we know little about intelligence or how to achieve it, but the press (mostly) ignores them. $\endgroup$
    – Dr. Snoopy
    Nov 23, 2023 at 6:46

1 Answer 1

0
$\begingroup$

I think there are a few reasons why the so-called leaders or experts of the AI field may continue to entertain the idea that creating an AGI is possible or is not impossible

  • the concept of an AGI is often not clearly defined in discussions - often, I'd assume that people think of an AGI as a program or machine which would have all capabilities of a human or even more, but probably not less - so you can't really do science - in other words, all these discussions are just philosophical and probably pointless
  • there's no clear evidence or proof that it's impossible to do it; anyway, almost no one is really trying to do this rigorously, i.e. from a mathematical/theoretical standpoint (an exception is the work on AIXI); I think we can't prove this until we clearly define the concept of an AGI, but other people may have a different idea, i.e. they try to develop narrow AIs, and, eventually, they think, something like an AGI (a human) will be developed, even if you don't really define it precisely/mathematically
  • people like to dream
  • it might attract investors and funding
  • most people don't really care about most other people

I've tried to give you a few possible explanations. I'd also like to note that there may be leaders/experts who try to sensitize the masses. Some, like LeCun (Turing Award), even claim there's no such thing as AGI (see slide 66 here), then he defines intelligence as a collection of skills, etc., in other words, he gives one definition of what we could call AGI. As you can see, everyone has their own opinion - and people are pursuing different things.

Having said this, I think your question can, in general, lead to opinions.

$\endgroup$
1
  • $\begingroup$ @NeilSlater What I write in my answer is not just an opinion. I can show you that many so-called leaders or experts are referring to different concepts when they mention AGI. What I write in my answer is a list of possible reasons why leaders/experts are not trying to clarify the real state of AGI development. The only real opinion I write is that I think we can't prove anything until we define clearly what an AGI is. This question would kind of fall under the social category of AI. Having said this, I actually voted to close the post. $\endgroup$
    – nbro
    Nov 21, 2023 at 12:56

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .