🤖 Yann LeCun, Meta — about AI dominance, OpenAI's closeness, and superintelligence
In a recent comprehensive interview with Wired, Yann LeCun, Chief AI Scientist at Meta, shared his insights. The scientist shared his predictions regarding the future of AI and explained his reservations about OpenAI. Here are the key highlights:
🔠 About human-level AI
Machine learning is great. But the idea that we're going to scale up the techniques that we have and get to human-level AI? No. We're missing something big to get machines to learn efficiently, as humans do. We don't know what it is yet.
🔠 About AI's role in society
AI actually mitigates dangers like hate speech. At Meta, we’ve had enormous progress using AI for these things. Five years ago, of all the hate speech that Facebook removed, AI preemptively took down about 25%. Last year, it was 95%.
🔠 About open source
Imagine a future in which the digital world is mediated by an AI system. You do not want that AI system to be controlled by a small number of companies on the West Coast of the US.
🔠 About OpenAI
When they started, they imagined creating a nonprofit to do AI research as a counterweight to bad guys like Google and Meta. But now, they are basically a contract research house for Microsoft. I think the research world doesn't care too much about OpenAI anymore.
🔠 About the superintelligence
There's no question that machines will eventually be smarter than humans. We should feel excited about this. The thing that excites me the most is working with people who are smarter than me, as it amplifies one's own abilities.
🔠 About dominance
There is no reason to believe that just because AI systems will want to dominate us. People are mistaken when they imagine AI systems will have the same motivations as humans. We'll design them not to.
💯 TIME magazine included LeCun in the list of the 100 Most Influential People in AI 2023.
📹 LeCun also discussed his vision for the future of AI on CBS News.
#Meta @hiaimediaen
