“Latest AI News, Apple Vision Pro & Bitcoin Surge w/ Salim Ismail” — Peter H. Diamandis Moonshots EP #87

Episode summary

Another “WTF just happened in tech” roundup with Diamandis and Salim Ismail covering: Apple Vision Pro launch (ecosystem integration is the real differentiator, not standalone hardware; B2B specialist applications like surgery/training are the killer use case); Sam Altman’s $7T chip raise (“half the GDP of the US”); Google’s Gemini launch and the existential search question; self-driving progress (Tesla replacing 300K lines of C++ with 3K lines of neural net code; Waymo car attacked in SF Chinatown); the Vesuvius Challenge ($700K prize using AI to decipher ancient scrolls); FCC banning AI-generated robocalls after Biden deep fake; Andrew Ng’s argument that more intelligence is inherently good; and Bitcoin’s continued surge. The most substantive thread is around deep fakes and “AI for truth” — both argue that deep fakes should be as illegal as counterfeiting, and that truth-verification AI needs to be embedded at information chokepoints (social platforms).

Key arguments / segments

• [00:02:00] Apple Vision Pro: ecosystem integration is the differentiator; B2B applications (surgery, training) will be the real killer use case; consumer XR headsets historically gather dust
• [00:09:00] Self-driving: Tesla replaced 300K lines C++ with 3K lines of neural net; Blackberry outage in Abu Dhabi caused 40% drop in accidents; Waymo attacked in SF Chinatown
• [00:17:00] Sam Altman’s $7T raise: “half the GDP of the US”; would make computation massive at tiny device level; silicon compute now exceeding 8B human brains’ worth
• [00:20:00] Google Gemini vs ChatGPT: Google’s search habits are sticky; moving send button by pixels drops usage dramatically; all will converge to single AI entry point
• [00:25:00] Vesuvius Challenge: $700K prize to read ancient scrolls with AI; discovered Epicurean philosopher throwing shade at Stoics
• [00:30:00] Deep fakes & AI for truth: Biden robocall; deep fakes should be illegal as counterfeiting; need AI truth verification at social platform chokepoints; truth-index for posters
• [00:38:00] Andrew Ng on AI risk: arguments for AI wiping out humanity are “so vague they boil down to it could happen”; more intelligence = more wisdom; AI can run billions of simulations

Notable claims

• Silicon compute on planet now exceeds 8 billion human brains’ worth of compute
• Tesla replaced 300,000 lines of C++ with 3,000 lines of neural net code for FSD
• Blackberry 3-day outage caused 40% accident rate drop in Abu Dhabi (distracted driving proxy)
• 259 selfie-related deaths reported in 2018 study
• Sam Altman’s $7T raise equals combined value of Apple + Microsoft + Google at the time

Bias / sponsor flags

• Fountain Life sponsorship: standard mid-roll by Diamandis
• Both hosts are Silicon Valley insiders with personal relationships to most companies discussed (Google, Tesla, OpenAI)
• OpenExo (Ismail’s company) promotion woven into discussion
• Abundance Summit promotion throughout (multiple speakers mentioned as upcoming guests)
• No bearish perspectives on any technology discussed; uniformly bullish framing

Relevance to Ray Data Co

Low-moderate. The “AI for truth” discussion and the concept of truth-verification at platform chokepoints is relevant to our thinking about data integrity. The observation that silicon compute now exceeds human brain compute is a useful framing milestone. The convergence thesis (AI + XR + crypto) is worth tracking but presented at headline level without depth.