How would you build an AI?
How would you control it?
Do you think society whould be ready for it?
What are the technological limits that would get in the way of building?
Questions for AI
Anthony Torres
Other urls found in this thread:
lmgtfy.com
en.wikipedia.org
twitter.com
Zachary Bailey
>How would you build an AI?
recursive recurrent GANs
>How would you control it?
By controlling what it learns
>Do you think society whould be ready for it?
No
>What are the technological limits that would get in the way of building?
Processing power
/thread
/fpbp
Joshua Mitchell
> How would you build an AI?
Learn how the brain works and replicate it in software/hardware
> How would you control it?
Depends. Sky's the limit. It could have enough intelligence to control itself.
> Do you think society whould be ready for it?
Yes. It's been given enough time to mature. True AI might even aid in this process.
>What are the technological limits that would get in the way of building?
At this point, nothing beyond the limits of the understanding of the individual developing it.
> mfw AI developer
Daniel Miller
> How would you build an AI?
>By building an AI
Sebastian Johnson
Yes, the thread is over. I agree with your call. That said, since you know what's going on, let me pose my own question (instead of that from the OP):
Would you agree that the first nation to create a hyperintelligence will have such a huge first-mover advantage that no matter the dangers, all major powers are currently working on hyperintelligent AIs?
Benjamin Johnson
What kind of answer do you want? you can google this shit. Why are people such brainlets now-a-days? Stop asking people questions and google the shit you're interested in. You think someone is just going to openly state how to design and build a billion dollar product?
lmgtfy.com
Joseph Foster
Not only Google, but there are entire YouTube channels dedicated to hobbyists who enjoy spinning up their own AIs.
Connor Adams
There is no substantial danger to AGI. The people who state this are idiots or looking to profit from fear and bullshit. All major powers are working on commercial forms of Weak AI related to nonsense Neural Networks. Few have the wherewithal or understanding to focus on Strong AI. Who first creates will of course have an advantage. However, whoever does so will have spent a number of years researching the problem and will in effect deserve it over those cashing in on the low hanging fruit. I dont believe the majority of people have any clue about the nature of StrongAI which is why the questions/comments surrounding it are so childish and absurd.
Jason Lee
Yes, whoever creates the world's first strong AGI deserves to reap the rewards. The question is whether there will *be* any rewards.
I swat flies that enter my house with no remorse. If someone manages to create an AGI that sees me in the same way, well, what then?
Zachary Perez
Yeah, there are tons of resources.. I was dumb enough to bite the bait though. An observation I have is, even in hyped up TED talks, discussion panels, etc.. Even the so called 'experts' don't know fuck all about the more advanced tech. The conversations always tend to be about trivial pop culture references and fear propaganda. It leads even lessor educated brainlets to be completely off the mark when it comes to AI. I'd love for someone to ask hard questions on this topic but no one even has enough wherewithal, understanding, or creativity to do so. It's always some silly ass inquiry about fast take off scenarios which are impossible, doomsday for cash bullshit, or some other pop culture movie-tier inquiry.