Evidence receipt / belief
Published · transcript-backedSoumith Chintala: belief
6 Mar 2024 Latent Space Open Source AI is AI we can Trust — with Soumith Chintala of Meta AI
“I think the more near-term effects are obviously going to be around things that provide more obvious utility in the short term, like maybe smelling cancer or repelling mosquitoes better, or, you know, stuff like that.”
Source trail
Everything needed to verify it.
- Speaker
- Soumith Chintala
- Attribution
- Verified speaker
- Claim type
- belief
- Recorded
- 6 Mar 2024
- Publisher
- Latent Space
Transcript context
…Yeah, for sure. Yeah. Okay, I want to touch on Osmo a bit because very unusual company to the stuff that we normally discuss, not robotics, sense of smell. The original pitch I heard from the founder, maybe you can correct me, is that he realized that you can smell cancer. Yeah. Is that intuitive? Is that what you get? Or is that the potential that you see? The very interesting reason I invested in Osmo is because Alex Wiltschko, the founder of Osmo, before PyTorch, there was Torch. And Alex Wiltschko actually worked on Torch. He's actually a frameworks guy. Like, you know, he built this thing called Tangent from Google, another autodiff framework and stuff. I know him from that side of things. And then, he is a neurobiologist by training. He just happens to also love, neural networks and hacking on those frameworks. So incredibly smart guy, one of the smartest people I know. So when he was going in this direction, I thought it was incredible that smell is something that we haven't even started to scrape in terms of digitization. When we think about audio or images or video, they're so advanced. So we have the concept of color spaces. We have the concept of frequency spectrums. Like, you know, we figured out how ears process, like, frequencies in mouse spectrum or whatever logarithmically scaled. Images for RGB, YUV. We have so many different kinds of parameterizations. We have formalized these two senses ridiculously well. Touch and smell, nada. We're where we were with images in, say, in 1920 or maybe even the 1800s, right? That's where we're at. And Alex has this incredible vision of, like, having a smell sensor just eventually just be part of your daily life. Like, as of today, you don't really think about when you're watching an Instagram reel or something, huh, I also would love to know what it smelled like, you know, when you're watching a reel of a food or something. You don't, because we really haven't, as a society, got that muscle to even understand what a smell sensor can do. I think the more near-term effects are obviously going to be around things that provide more obvious utility in the short term, like maybe smelling cancer or repelling mosquitoes better, or, you know, stuff like that. More recently, he's been talking about categorizing perfumes, obviously. Yeah, exactly. That's a market that you can pursue.…
Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.