David Bau on How Artificial Intelligence Works

Historically, as an engineer, as a computer scientist, it's really been our responsibility to understand the systems that we make, to make sure that they're doing what we want, that they operate correctly. And and I feel that the new discipline of machine learning, because it's become so important to just accept these black boxes and use them even though we don't understand them, is leading to really an unhealthy turn. in in the uh in the practice of engineering in computer science, uh, you were training a whole new generation of computer scientists to be comfortable with this idea that they shouldn't really look inside these complicated black boxes. And now the good fight. With Yasha Monk.

There is at the moment, for good reason, a lot of debate about artificial intelligence. How capable are current models? Are we approaching something like artificial general intelligence? And how big are the impacts on the world going to be? Is this, for example, going to lead to a massive loss for jobs, not only for drivers and other kinds of blue-collar professions, but also perhaps for a lot of white-collar professionals?

Well, one thing that strikes me about this debate is that it's often held without real knowledge of the underlying technology. And so I was really intrigued to learn more about the nuts and bolts of how AI actually works. And when I was recently at a conference about artificial intelligence at Harvard, I had the good fortune of running into David Bao. David was at Google for a long time, is now a computer scientist at Northeastern University, and he is really good at explaining what a neural network is, how it is that you train an eye model and what that means for how to think about this technology.

We already have this episode, so unlocking it is cheap and permanent: the full transcript (56 minutes, 9,186 words), segment and word-level timestamps, and .txt, .srt, .vtt and word-level JSON downloads, as many times as you like, forever.

No account needed, we create one from your checkout email. With credits it is 1 credit; packs start at 5 episodes for $3.49, and monthly plans are cheaper still.