AI Global Odyssey: REAL Situational Awareness — Manifold #63
AI in light of recent travels to SF, Singapore, Manila, Berkeley, and Silicon Valley. See A Month on the Road for photos and more discussion of some of the travel.
Chapters:
(00:00) - Overview: SF, Singapore, Manila, Berkeley
(01:48) - The AI Bubble in Silicon Valley
(04:00) - Scaling Laws and AGI
(23:36) - Global AI: Singapore, Philippines, real Enterprise applications
(34:59) - AGI: Manhattan Project? Manifest and P(doomers), Situational Awareness
(51:00) - China LLMs, Huawei vs Nvidia GPUs, US vs China AI race
Links:
Scaling Laws for Neural Language Models: https://arxiv.org/pdf/2001.08361
AI rollout in Philippines Call Centers:
LLM rankings and Qwen2: https://huggingface.co/spaces/open-llm-leaderboard/blog
The Economist on China LLMs: https://archive.ph/nW7ch
Situational Awareness summary: https://x.com/hsu_steve/status/1803414701159714825
Audio-only version and transcript:
https://www.manifold1.com/episodes/ai-global-odyssey-real-situational-awareness-63
Really valuable update to stay clued in on what’s happening.
Great to see the DJI pocket in use! :)
You may be right to identify this as a bubble. One thing that Tyler Cowen is fond of pointing out is that it's harder than you think to identify actual financial bubbles in history - the late 90s/early 00s internet boom was not wrong to assign a lot of value to tech companies, just a bit early.
Likewise people were calling the 08 house prices a bubble, but in retrospect it doesn't look like a massive deviation from the long term trend: https://fred.stlouisfed.org/series/CSUSHPINSA
So the lesson seems to be that financial markets usually are predictive of long term valuation creation but sometimes overoptimistic about the timing. I think that this matches your own belief - AI will indeed be a game changer, but it's going to take longer than the SF crowd think to show up in earnings & macroeconomic data.