ฝัง
- เผยแพร่เมื่อ 7 พ.ค. 2024
- The more human-like artificial intelligence becomes, the more we understand about how our brains actually work. Through that discovery process, researchers are identifying ways to design artificial intelligence in ways that factor in the safety and morality of their potential impact.
Greylock general partner Reid Hoffman interviews Dr. Fei-Fei Li, the co-director of Stanford's Institute for Human-Centered AI, and Mira Murati, the CTO of OpenAI. In this interview, they discuss how technology like GPT-3 is being trained with human safety in mind, and how academia, industry, and policymakers are coming together to ensure AI is developed and deployed in ways that benefit all.
This interview took place during Greylock's Intelligent Future event, a daylong summit featuring experts and entrepreneurs working in artificial intelligence.
You can read the transcript from this interview here: greylock.com/greymatter/ais-h...
Chapters
0:00 Introduction
1:00 Redefining a North Star
3:10 AI Paradigm
5:18 Creating Safety Norms
8:35 Human-Centered AI
11:45 Beyond the Model
15:07 AI Research Ethics
19:24 National Research Cloud
21:48 Amplify Human Creativity
27:17 Real World Effects
29:23 Innovation Versus Regulation - วิทยาศาสตร์และเทคโนโลยี
Balkan is proud of you Mira ❤ 🇦🇱🇽🇰
EXCELLENT ideas! THANK you so much for sharing!
Always enjoying hearing from Dr. Fei-Fei Li and CTO Mira Murati. Love the learning. Nice choice of words with good intentions "beneficial" "infusing ethical and human centered values." "Replace or augment". Lots of points to consider in safety and security of AI.
What are the recent contributions of openai to world of open research? Beyond papers most models aren't released and code kept unpublished. They did publish spinning up rl back in the day, but now deepmind delivers more cool stuff like open sourcing mujoco. I'm talking robotics.
GPT-3 is really a serendipity engine, more than anything else, doesn’t matter if it sometimes comes up with whacky answers, as long as it makes you think anew.
23:09 wow, that’s really something, and precisely the power of the thing that I am talking about…
Isn't that also modern art? The artist who throws paint at a canvas or poet John Cage putting random phrases together for a lecture.
Very good.
Thank you for this interview! Can I suggest that you add audio to the intro and/or make a bit shorter?
Agreed. I thought my speakers were broken.
So many windows opening and at the same time, very challenging implications hidden in the months to come. I wouldn't´t ask for another era to live in... :-)
Wonderful talk even though safety was not addressed in any significant way. Nonetheless both interviewees emit brilliance in the domain of AI. When they got to discuss safety, it was more about ethical behavior. They seem to imply that AI needs guardrails to ensure continued safe and ethical behavior in this realm of AI. But who decides what is ethical? If we democratize that decision making process, does the AI entity participate? Do we need to inject the concept of religion into AI so that there is the fear of consequences for poor and unsafe behavior? Can you see HAL going to confession - joking aside would some sort of religious influence have inhibited HAL from killing humans? Anyhow, that to me are the long-term mitigation efforts needed to address security concerns associated with AI. At the end of the day is AI going to further influence a Stratosian and Troglite divide? A lot of philosophical elements here.
Good comment!
A.I and biotechnology are the future
robotics is the way to get us into agi.
Wow this lady is a genius
I agree. What do you think about AI with metaverse. There’s a project named Alethea AI. They are fusing AI with nfts to make them intelligent in metaverse enabling them to talk and have trainable personality
Interesting. Not sure how Alethea's efforts advance humanity though.
I follow Stanford university courses and Andrew gn.
Vive me code to use
Respect for a woman, her work is very huge, but my heart is breaking from such injustice, why is there little talk about her? Do a lot of interviews with her, I want to see more women and girls in the IT sphere !!! ❤
Hi.
mIrA murAtI
You can tell these 2 ladies never talked to each other before. What I'm saying is: It's a battle between all that involved. There's not much collaboration between firms. Whoever wins out will dominate the AI world. I placed my bet on Tesla. Let's see if it works out.
MSFT 🎉
Female leaders 🎉
Uhim.
The most useless conversation on AI
you've ever seen.
Yes, nice people.
No, zero value.
You have spoken my mind. Fei-Fei Li is a bunch of bullshit. Overrated. Way-overrated.
Leave something better, then.
@@geirasES The creator of THIS video asks for a comment.
That's why he allowed the comment function, Geiras.
Fei Fei is the real deal you just a hater. Really understands her field in depth and makes it easy to understand for non-academics with real world examples without trying to sound too smart. Watch some of her lectures in machine learning instead of just talking nonsense
@@thefamousdjx All their lectures are excellent.
This one is not. Blah blah.
Unless she was hired to lecture in this manner.
Mira does not know what she is talking about
what make you think that "Mira does not know what she is talking about"
@@xtu373 because he has Slavic Balkan origin they hate Albanians but love our Land
Not many know better than her. Musk hired her to head 2 major groups at Tesla. Open AI appointed her position of CTO because of her combined math, computing and advanced engineering competence.
@@zenco1611 from where you find this?