Interview w/ John Carmack on Lex Fridman podcast
Interesting segments:
AGI
John is one of a growing number of people that think AGI is not only achievable, but has a 50% chance of being achievable near the tail end of this decade 2030-2032. His personal take was that we'd see hints of it by 2030, AIs that were equivalent to in his words a "learning disabled toddler". He thought this would be sufficient to kick start the next phase of innovation. He made the point (which I agree with) that consciousness is not substrate dependent - meaning it can be born out of technology just like biology. Interestingly, John has taken a step back from his work at Meta (he was Oculus CTO and stayed on for a while) to focus on solving AGI.

I'm thrilled about the growing number of people are working on this, with Deep Mind being the highest profile. Facebook is doing similar work that we hear about in drips like this: https://www.infoq.com/news/2020/09/facebook-ai-genomics/

Mars:
If SpaceX IPOs and/or if Elon ever stopped being the CEO, any Mars colony is probably off the table. A profit driven shareholder owned business would never end up investing enough time and resources to get to Mars (at least not for a while). He thinks we'll probably have some equipment there by 2030 but not people. Obviously, if Elon maintains control through an IPO and stays on board then Mars is still on the table. John is a long time friend of Elon's and says he is dead set on Mars, but without him it's too expensive a proposition for any other group/agency/etc to actually tackle.

Metaverse.
John pointed out something that I've long thought, which is that it is not possible to spend as much money as Meta is efficiently. He said when he saw how much they were spending it made him sick to his stomach. He thinks Meta could be 2-4X (magnitude) more efficient when looking at how effective smaller non-Meta teams are able to work on similar technology. Meta's size, bureaucracy, need for intra-team compatibility, and need to do things carefully for PR reasons slow it down dramatically. He thinks one of the primary thing Meta gains from spending so much is communicating how serious they are, which helps attract talent.
No comments yetBe the first to add your insight!

Author

Related