World's Best Dota 2 Players Beaten by AI

Discussion in 'General Offtopic' started by Roboartist, Aug 12, 2017.

Share This Page

  1. Roboartist

    Roboartist Reclaimer Moderator

    Messages:
    4,003
    Likes Received:
    1,119
    Trophy Points:
    123


    Really interesting watch
     
    • Like Like x 3
  2. NRG Conq

    NRG Conq Former Owner Founder

    Messages:
    5,167
    Likes Received:
    1,229
    Trophy Points:
    123
    This is just insane, the guys like "do you think he has a shot" dude just sounds like he lost faith in humanity when he says "No"
     
    • Like Like x 1
  3. Idiot

    Idiot Vitor Belfort '97 hand speed

    Messages:
    888
    Likes Received:
    537
    Trophy Points:
    93
    • Like Like x 1
  4. OP
    Roboartist

    Roboartist Reclaimer Moderator

    Messages:
    4,003
    Likes Received:
    1,119
    Trophy Points:
    123
  5. Bestworld

    Bestworld MBA

    Messages:
    1,877
    Likes Received:
    883
    Trophy Points:
    123
    AI is awesome. I would love to get into AI development one day. It's the future for sure.
     
  6. NRG Conq

    NRG Conq Former Owner Founder

    Messages:
    5,167
    Likes Received:
    1,229
    Trophy Points:
    123
    ye but m8 its like using cheat codes.

    I'm kidding, maybe, do we really think we can control it no matter how advanced AI can become, and if this kind of technology becomes widely available some day than what does that imply for our chances of doing so? Perhaps we're not ready for this as a species =P

    This is a complicated philosophical issue, there are some who think everyone should have AI capacity, and perhaps they should but more importantly can we find a way to only have the good AI? Maybe that is not necessary and if we're lucky we will get to the point where there is defensive and aggressive AI and that can basically allow us to remain safe by nullifying each other but with the spike in cyber warfare, rampant growth in artificial intelligence, and global instability because of war, and the chances of global warming playing a role, as well as nuclear warfare, it's not so safe of a bet. Perhaps AI can be our saving grace, but if it is so than we need to find a way to make it safe for us as well because if its not it poses a more dangerous threat than anything, in short everyone will have to have AI level protection and possibly quantum encryption, or no one can, or perhaps just the extremely powerful will have these technologies.
     
    Last edited: Aug 12, 2017
  7. OP
    Roboartist

    Roboartist Reclaimer Moderator

    Messages:
    4,003
    Likes Received:
    1,119
    Trophy Points:
    123
    Just be sure to do it soon before they start programming themselves
     
  8. Bestworld

    Bestworld MBA

    Messages:
    1,877
    Likes Received:
    883
    Trophy Points:
    123
    I honestly just wanna see if the world is capable of transitioning into a state of general income, which is needed once nearly every sector is automated. Probably won't be in my lifetime, but oh well, I can dream.

    I've always been drawn towards AI. I even did a FUCKTON of research on the Enigma machine back in highschool, just before going into uni. Hopefully I will become skilled enough to work with it one day.

    #HalWillBeReal
     
    • Like Like x 2
  9. NRG Conq

    NRG Conq Former Owner Founder

    Messages:
    5,167
    Likes Received:
    1,229
    Trophy Points:
    123
    I've felt the same way, and i'm no programmer but I totally agree, AI is risky though, is it worth the risk, it's sort of an all or nothing kind of thing eh? The more important question is if that risk has already been taken, it's probably to late to stop the development of AI.
     
  10. Bestworld

    Bestworld MBA

    Messages:
    1,877
    Likes Received:
    883
    Trophy Points:
    123
    That's one of the many concerns of AI actually.
     
  11. NRG Conq

    NRG Conq Former Owner Founder

    Messages:
    5,167
    Likes Received:
    1,229
    Trophy Points:
    123
    They probably are in some corporate or government lab.
     
  12. Bestworld

    Bestworld MBA

    Messages:
    1,877
    Likes Received:
    883
    Trophy Points:
    123
    Yes, the risk is great, especially if you wanna obtain true AI. There should just be a rule against making any combat or war-related AI systems, but that's already too late, so....

    As to whether it's worth the risk or not, that depends how far you wanna take AI. You can add limitations. I take the nutty scientists stance on this one though. LET THE AI BE FREE!!
     
  13. Bestworld

    Bestworld MBA

    Messages:
    1,877
    Likes Received:
    883
    Trophy Points:
    123
    Depends what you mean by "programming themselves". What you saw in this vid was also an AI basically "programming itself" from playing itself. So in a sense, it's the ideal training environment for them to "program themselves" in this way. It depends on the type of AI you want to develop though. My concern still lies with military AI as it's already been made and being improved.
     
  14. Hashbrowns

    Hashbrowns Veteran

    Messages:
    2,465
    Likes Received:
    502
    Trophy Points:
    113
    Last edited: Aug 12, 2017
    • Like Like x 1
  15. OP
    Roboartist

    Roboartist Reclaimer Moderator

    Messages:
    4,003
    Likes Received:
    1,119
    Trophy Points:
    123
    • Like Like x 1
  16. Hashbrowns

    Hashbrowns Veteran

    Messages:
    2,465
    Likes Received:
    502
    Trophy Points:
    113
    Did they get shut down because they were dangerous, no?

    Did they invent their own language, yes.

    Probably not good.
     
  17. OP
    Roboartist

    Roboartist Reclaimer Moderator

    Messages:
    4,003
    Likes Received:
    1,119
    Trophy Points:
    123
    There were other reports indicating that the reason was because Facebook didn't incentivize using English when set to talk with one another. But all things considered, it's still impressive and/or maybe a little unsettling. I just don't know enough details on the matter.
     
  18. Leek

    Leek Veteran

    Messages:
    2,332
    Likes Received:
    226
    Trophy Points:
    63
  19. Selaya

    Selaya Ɛ> Head Admin

    Messages:
    866
    Likes Received:
    473
    Trophy Points:
    63
    that wont work, unless you create a non-deterministic computer because determinism and creativity are like, polar opposites
     
    • Like Like x 1
  20. OP
    Roboartist

    Roboartist Reclaimer Moderator

    Messages:
    4,003
    Likes Received:
    1,119
    Trophy Points:
    123
    It all depends on whether or not a goal is set for it. If that goal is to discover a solution to an unsolved physics equation, or scale a wall too steep for it to physically be capable without augmentation, we provide the tools or information for it to solve. The ai isn't built for any of these tasks specifically, but is given the goal to solve regardless, much like the ai in the video.

    Deterministic reasoning is fallible when talking about advanced ai because (which I'm sure you know), to an extent, we're no different. We are nothing more than biological machines, whereas an ai is synthetic. We are usually not deterministic, as I had the choice to even make this thread. Now currently for general machine learning, ai is dependent on an external force to provide a goal. But we are our own evidence that the possibility of an independent ai is possible. There's no divine reasoning in the realm of science to prove otherwise.

    Once curiosity is successfully implemented into the ai's system, there's no telling what it would do to solve a problem. Curiosity (apart from need) is the prime incentive for discovery.
     

Users Who Have Read This Thread (Total: 0)

Share This Page