The Dark Side of Competition in AI | Liv Boeree | TED

144,904
0
Published 2023-11-09
Competition is a core part of human nature, and it can drive us to extraordinary feats. But when it goes wrong, the results can be devastating. Poker champion and science communicator Liv Boeree introduces us to "Moloch's trap" — the dark force of game theory driving many of humanity's biggest social problems, which is now threatening to derail the AI industry.

If you love watching TED Talks like this one, become a TED Member to support our mission of spreading ideas: ted.com/membership

Follow TED!
Twitter: twitter.com/TEDTalks
Instagram: www.instagram.com/ted
Facebook: facebook.com/TED
LinkedIn: www.linkedin.com/company/ted-conferences
TikTok: www.tiktok.com/@tedtoks

The TED Talks channel features talks, performances and original series from the world's leading thinkers and doers. Subscribe to our channel for videos on Technology, Entertainment and Design — plus science, business, global issues, the arts and more. Visit TED.com/ to get our entire library of TED Talks, transcripts, translations, personalized talk recommendations and more.

Watch more: go.ted.com/livboeree23

   • The Dark Side of Competition in AI | ...  

TED's videos may be used for non-commercial purposes under a Creative Commons License, Attribution–Non Commercial–No Derivatives (or the CC BY – NC – ND 4.0 International) and in accordance with our TED Talks Usage Policy: www.ted.com/about/our-organization/our-policies-te…. For more information on using TED for commercial purposes (e.g. employee learning, in a film or online course), please submit a Media Request at media-requests.ted.com/

#TED #TEDTalks #ai

All Comments (21)
  • @brandontylerburt
    In the film "Metropolis" (1927), overcome by the hubris and soulless destruction wrought by the underground machinery, the main character exclaims a single word: "Moloch!" Surely the image it invoked for 1927 audiences was vivid and immediately accessible, but I had only a vague idea what Fritz Lang meant by it—and here, Liv Boeree coins such an apt metaphor for a kind of trap that, despite its distressingly familiar omnipresence, is still not easy for humans to detect or understand as we go about our business. If the metaphor catches on, it will go a long way toward helping us avoid many existential disasters.
  • @SeanLumly
    Why is that academia is NEVER mentioned in the same breath as innovation, where the lions share of breakthrough innovations and the training of innovators happen? Corporations are risk averse -- they "innovate" only if there are short-term profits to be made, and these tend to be safe. More substantial innovation are often aligned with universities and research institutions that have significant public funding.
  • @dsoprano13
    This is a well needed speech. Probably the most relevant one I've seen this year. Just like a sport, you need a referee to step in and regulate.
  • @NathanHarrison7
    Excellent presentation. In short, she’s asking the worlds governments and the worlds largest. and most powerful, companies to put people before profits. Makes sense, but it will never happen in this world.
  • @NathanRennard
    I'm in accord for changing our mindset from scarcity to abundance. The choice here will determine whether we continue to compete (mindset of scarcity) or change to a mindset of potential abundance, and begin cooperation and contribute towards a harmonious alignment.
  • @chuchuman1163
    I really like the "Moluch's game" analogy that Liv Boeree uses to explain the dark side of competion in the AI industry. I give this video and Ted Talk a thumbs-up imo. 👍
  • @alexachtar9580
    Liv the fact you are calling out the Moloch agenda has exceeded my respect for you. My favorite poker player of all time and now one of my favorite humans. Thank you for promoting AI safety in a dangerous world, in san fransisco wish i saw this live. Much love
  • @jeremyconner8022
    Fantastic job. Absolutely hit the nail on the head when you talk about the problem being the incentives, not the players. Humans are, always have been, always will be, fallible. If you want to predict the future of any process, evaluate the incentives in that process. If you want to direct the future in a more positive way that promotes a balanced ecosystem, the defining incentives are the only thing that gets you there. Loved this TED talk. Well done.
  • @littlebluefishy
    Thanks for all these people who are recognising the problem and is willing to stand out and talk about it. Sacrificing own prospect for the bigger good is certainly not easy, but it is a needed effort. Thanks all the anonymous heroes that are going to take part in helping to stop molic.
  • @AdityaMehendale
    "Moloch's trap" - that term is going to catch-on. Well played.
  • @RT-eb6vo
    Brilliant. Thanks so much for putting this together and giving us all something new to think about.
  • @stchaltin
    I believe we’ve already reached AGI by a very important metric. It’s cheaper to train computers to perform tasks than it is to employ and educate humans to do the same. This is a fundamental problem. We have to rethink the entire concept of money, because our labor loses more and more purchasing power by the second.
  • @saranbhatia8809
    Wow what great talk and the message....thanks for this!!!💯
  • @Silent1Majority
    This was an absolute awesome breakdown on various issues, and an excellent offer for solutions.
  • @Ant3_14
    Well, role of CEO is to make company profitable as much as possible. We need to change incentives (for example require companies make actions to prevent harm or getting fine). It's naive to just count on good will of actors if incentives and rules of game result in different behavior.
  • @SanzAlfredo
    Qué claro que la tiene esta dama.! Me impresionó lo de los medios, porque acá en mi país estamos atrapados por una polarización cada vez más profunda. Saludos desde La Ciudad el Humo.
  • This is awesome and thank you for a viewpoint from this perspective, and love how you use AI to make those pictures :) Isn't it strange that AI art is already easily identifiable like it has its own watermark already.