The general consensus on AI for the last four years has been: we have finally hit the threshold above which we cannot fit what is necessary to a curve and this moment will last forever. It has never been correct.
I can feel it in the air right now with the adoption of Claude Code. All sorts of skeptics acquiescing to the idea, "well, OK, I guess we will be working like this forever, we'll all be architects, I accept this."
We may only be working like this for 4 months.
@kyle The thing is that, unlike hand crafting ascii, architectural guidance scales up indefinitely, with each rung on the ladder gaining combinatorially higher impact and complexity. I think these systems will almost certainly climb that ladder quite a bit, as they already have, but humans can and will always be one rung above until we have something that can climb all the way to the top (think: CEO).
Such a system would be, by definition, AGI.
This is a transcript of Lex Fridman Podcast #494 with Jensen Huang. The timestamps in the transcript are clickable links that take you directly to that point in the main video. Please note that the transcript is human generated, and may have errors. Here are some useful links: Go back to this episode’s main page Watch the full YouTube version of the podcast Table of Contents Here are the loose “chapters” in the conversation. Click link to jump approximately to that part in the transcript: 0:00 – Introduction 0:33 – Extreme co-design and rack-scale engineering 3:18 – How Jensen runs
@djspiewak 01:56:31 https://youtube.com/watch?v=vif8NQcjVf0&t=6991
The framing of the question was, effectively, a CEO capable of running a billion-dollar company, which made me remember this conversation. And reminded me that I never replied, which is unfortunate, because you gave me a great framing with the ladder analogy that I have often used since. So, here is a belated thank you.
Kyle Hughes
Daniel Spiewak