CEOInterviews.AI
Start App
Get an email when Mira Murati gives a new interview
One email per new interview. Unsubscribe in one click.
Mira Murati
Co-Founder & Chief Executive Officer, Thinking Machines Lab

AI's Human Factor | Stanford's Dr. Fei-Fei Li and OpenAI's Mira Murati on AI Safety

📅 Dec 17, 2024 Greylock 33 MIN 77 VIEWS 26 SEGMENTS · 5 SPEAKERS
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...

Questions asked in this interview

4
  1. 5:12So Mira, why don't we start with you just so we mix it up a little bit?
  2. 9:19And then the decision-making using the algorithm, the inference, how does it assist the humans or inform the humans?
  3. 12:29How do you take that and also help lead the way in thinking about this safety within the industry?
  4. 28:03And probably related to that is, Fei-Fei, you mentioned governing boards as an oversight, and I'm wondering how do you address the tension between the process instruction and the pace of innovation?
Heather Mack 0:04 ↗
Hi everyone, welcome to Grey Matter, the podcast from Greylock where we share stories from company builders and business leaders. I'm Heather Mack, head of editorial at Greylock. Today we're discussing how safety and ethical considerations must be factored into the development and deployment of artificial intelligence technology. Greylock General Partner Reid Hoffman interviews Stanford Professor Dr. Fei-Fei Li, who is the co-director of the Center for Human-Centered AI, and Mira Murati, who is the chief technology officer of OpenAI. This interview took place during Greylock's Intelligent Future event, a daylong summit featuring some of the leading experts and entrepreneurs working in artificial intelligence. You can watch the video of this interview on our YouTube channel and you can read a transcript on our website greylock.org.
Reid Hoffman 1:00 ↗
Universe, because both Fei-Fei and Mira are accomplished technologists massively beyond the scope of safety, both building amazing things and have been part of historic contributions. I thought we'd start with a kind of more broad question, which is: what's currently most exciting you in AI? Right, Fei-Fei, I'll start with you, Mira, I'll go to you, and then we will dive into safety.
Fei-Fei Li 1:35 ↗
Wow, so many things are exciting me. But I guess I'll just say something that's on my mind right now. We are finishing the review process of a paper that is coming out which is redefining a North star for robotics. I feel like this moment working in robotic learning feels a bit like back in the days of ImageNet, where we're really imagining what can be done that would make robotics dream come true. And this particular paper that we are, hopefully fingers crossed, we're in the rebuttal review stage, will lay out a benchmark of a thousand robotic tasks that are inspired by actual human activities. So it's a scale we've never seen in robotic research.
Reid Hoffman 2:37 ↗
Can it be too much of a presage to say what one or two of the most interesting of those tasks are?
Fei-Fei Li 2:51 ↗
Yes. All of the thousand tasks come from the American Labor Bureau survey of American time usage and their equivalent of European government agency, where we look at what humans do in their daily activity. And we actually did quite a bit of study on what people want robots to help with. For example, most people do not want a robot to open your Christmas gift, apparently we still want to do that ourselves. But almost everybody wants the robot to clean a toilet, so that's ranked really high. Packing kids' lunch is actually packed ranked fairly high.
Reid Hoffman 3:39 ↗
Interesting. So we're talking about real human activities instead of how in robotics we tend to see toy examples.
Fei-Fei Li 3:47 ↗
Yes, exactly. This is going to be a thousand real tasks. Awesome.
Mira Murati 3:54 ↗
Yeah. What's most exciting to me these days is seeing how far we can push the paradigm that's driving all AI development in the field, which is this combination of large neural networks with a ton of data and a vast amount of compute. In the past few years, we've seen this formula drive a ton of progress in AI research. We saw that with GPT-3 and then again with Codex and the first generation of DALL-E. So from OpenAI's perspective, we're trying to build these general systems that can think of the world in a similar way that humans do, systems that have a kind of robust concept of the world. So if an AI system sees an avocado or reads the word avocado, the concept that gets triggered is exactly the same. And it's exciting to see that we've got systems that have achieved some sort of linguistic competency and understanding of visual concepts, and we'll continue to push this paradigm ahead.
Reid Hoffman 5:12 ↗
So let's shift to safety, because both of your organizations in different ways are some of the most important industry leaders on what we encounter with thinking about AI safety. So Mira, why don't we start with you just so we mix it up a little bit? Say a little bit about OpenAI's approach, what kind of being this novel organization organized around a 501(c)(3) and organizations like this, how you're trying to identify what safety is and how to create those norms both in your own action but also catalyzing other industry.
Mira Murati 6:02 ↗
As you know, our mission is to build a general system and figure out how to deploy that beneficially to the world. You know, just one word 'beneficially', but figuring out how to robustly do that is actually an immense challenge. It's hard to predict the future. It is hard to predict all the ways in which these systems might create harmful biases or other risks that we can't even imagine. But at the very least, we can try to get as much understanding as possible, gather as much knowledge as possible, and leave the options open. And that's OpenAI's strategy. We are trying to deploy these systems continuously but in a controlled way. That means an API. So GPT-3 was first deployed through an API to a small group of users, and then eventually we broadened access as we understood how to get a good handle on the risks. But I think it's really hard, and that's actually part of the reason why we decided to deploy GPT-3. Because if you are in the research lab, you may think this is going to be the most prominent risk based on what we see, but when the model comes into contact with the real world, that gets tested. And we found out several times that we deployed that we were wrong and the most prominent risk was something else. For example, with GPT-3, we were convinced that misinformation was going to be the most important risk, and it is very important, but in practice we saw that spam was actually a much bigger risk that we had to focus on. And the same thing happened with DALL-E as well. So I think it's really important for these models to come in contact with real users, with the real world, and understand where the friction is, where the limitations are, and iteratively build mitigations. And the mitigations we're building are probably not future-proof, but it's a place to start and it gives us enough knowledge of where to go. But at the same time, we need to think about how the complexity increases as the models become more and more capable. So for example, with language models now, we oversee the output of the model by a human for sensitive use cases, and it's not something that would scale with more powerful and advanced models. So then we have to come up with techniques to help humans evaluate the output of these models. And OpenAI has been working with other language model developers to coordinate on some of these standard practices to figure out how to deploy language models safely.
Reid Hoffman 9:08 ↗
Fei-Fei, same question, but as opposed to OpenAI, Human-Centered Artificial Intelligence Institute, role of universities with industry, governments, plus the work you guys are doing.
Fei-Fei Li 9:19 ↗
Great question, Reid. Safety is one of those words like health. Everybody wants it but it's really hard to define it. So we, the co-director of Stanford Human-Centered AI Institute, which we call Stanford HAI, think a lot about what we really want for AI for future AI, and we come to the word human-centeredness. We focus on infusing the human-centeredness into every stage of AI research, development, education, policy work. Obviously we're a university, so we don't deploy products, but we hope that what we inspire as applications will have an impact in productization downstream. So AI is not one thing. Designing an AI system is really stages of work decisions, and we believe that every stage of this AI development we need to infuse the ethics and human-centered values into this. The simplest way to put it: how do we define a problem? For example, is your goal to replace humans without consideration of all the social implications, or augment human capability? That is before you write a single line of code, you already are thinking about human values. The data: where does it come from? How do you collect it? How do you ensure data integrity? How do you annotate it? There are a whole bunch of issues and considerations from fairness to privacy. Then the algorithm itself: how is it safe, is it secure, is it biased? And then the decision-making using the algorithm, the inference, how does it assist the humans or inform the humans? So every stage of AI development needs human consideration, and Stanford HAI is really trying to embody that process. And in this, it's not even a side product, a central product of that process is the people we educate: the students, the undergrads, the graduate students. When they come out of this, after years of working or learning from Stanford HAI's courses and the lab work, they become technologists or business leaders or policy thinkers who understand the human-centeredness of AI.
Reid Hoffman 12:29 ↗
So let's now dive a little deeper in each specific organization. So for DALL-E and GPT-3, the front end of the API is in part to understand what the possible risk flows look like so that you can begin to train those. How do you take that information and iterate to a beyond-human safety model? How do you take that and also help lead the way in thinking about this safety within the industry?

13 more exchanges in this transcript

Sign in free to read the rest of this interview. No card required.

Sign in to read the full transcript

Cite this transcript

APA, MLA, BibTeX
APA

Murati, M. (2024, December 17). AI's Human Factor | Stanford's Dr. Fei-Fei Li and OpenAI's Mira Murati on AI Safety [Interview transcript]. Greylock. CEOInterviews.AI. https://ceointerviews.ai/interview/567022/

MLA

Mira Murati. "AI's Human Factor | Stanford's Dr. Fei-Fei Li and OpenAI's Mira Murati on AI Safety." Greylock, 17 Dec. 2024. Transcript, CEOInterviews.AI, https://ceointerviews.ai/interview/567022/.

BibTeX
@misc{murati2024_567022,
  author       = {Mira Murati},
  title        = {AI's Human Factor | Stanford's Dr. Fei-Fei Li and OpenAI's Mira Murati on AI Safety},
  howpublished = {Interview transcript, Greylock. CEOInterviews.AI},
  year         = {2024},
  month        = {dec},
  url          = {https://ceointerviews.ai/interview/567022/},
  note         = {Speaker-attributed transcript with timestamps}
}