Fei Fei Li

speaker

223 appearances 1 recordings 1 series first heard Dec 2024 last heard Dec 2024

Fei Fei Li’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months

No recordings in the last 12 months.Older appearances are listed below; set an alert to hear about the next one.

Appearances

newest first · ▶ plays the moment

Armchair Expert with Dax Shepard · Fei Fei Li (on a human-centered approach to AI) · 11 Dec 2024 podcast

Here's the The difference between AI and nuclear technology is AI is so much more an enabler of so many good things. True. So that's very different from nuclear. Of course, nuclear can be an energy.

Right. But AI can help discover drugs. AI can help break through infusion. AI can personalize education. AI can help farmers. AI can map out biodiversity. So AI is much more like electricity than it is like nuclear physics. So that's the difference.

So from that point of view, the viewing angle of AI, at least I do not think it has to only from the competitive lens because it should be also through the enabling lens, the enabling of our world, of so many good things that can happen. And that's the challenge is how do we keep the dark use of AI at bay? How do we create that kind of

balance somehow, but in the meantime, encourage the development of AI so that we can do so many things that's good for people.

I do think we must always recognize cooperation is one of the solutions.

I do think we should pay a lot of attention on where rubber meets the road. Because AI can sound very fancy, but at the end of the day, it impacts people. So if you use AI through medicine, then there is a regulatory framework. For example, my mom, again, does imaging all the time because the doctors have to use MRI, ultrasound, you name it, to monitor her.

Honestly, do I care if that MRI is fully automatic or is it operated by humans or it's a mixture? As a patient family, I probably care more about the outcome. If the result of the MRI can be so accurate,

Exactly. But all I care are two things. One is it is the best result my mom can get. Second is it's safe, right? I don't care if it's that kind of mixture. So that regulatory framework is there. I'm not saying FDA is perfect, but it is a functioning regulatory framework. So if there is an AI product that goes into the MRI, I would like it to be subject to the regulatory framework. There we go.

Yeah, yeah. Right. So that's where rubber meets the road. The same as finance, environment, transportation, you name it. That's a very pragmatic approach. It's also urgent because as we have AI products that's entering our customers' market, And it takes away from, in my opinion, the science fiction rhetoric about existential crisis, machine overlord. That can stay with Hollywood.

I believe the downstream application is where we should put our guardrail attention at.

Thank you.

That's very sweet. That's the problem of always chasing after North Star. I try to like look forward. One thing I do reflect back is how grateful I am. I'm not here by myself. I'm here because of the Bob Sabella, Gene Sabella, the advisors, the students, the colleagues. That I feel very, very lucky.

Yeah, it's good.

A little bit. I didn't finish the audio part.

I'm like, do I have time? I should finish my Von Neumann book.

Yeah, you got a couple new books to read. I'm so great. You like the book.

I felt a little self-conscious because of my accent.

Because I consider if I should narrate my own book, but I feel like my accent is probably too strong for that.

Right.

Right. And that's how many hundred pages? You also probably need to put your time there.

Showing 201–220 of 223 · page 11 of 12