Technical AI moats & backup plans using SLMs for LLM price hike or collapse

Using a Prof G clip as the anchor, Goju maps the current AI capex-and-valuation moment onto 1999. Galloway's point is that the enormous infrastructure spending now rests on demand from essentially two players, that Meta is already leasing out capacity it overbuilt, and that this rhymes with the dot-com run-up. Goju layers his own reading on top: the reason he doubts the frontier labs are the untouchable giants people assume is that the transformer came out of Google Brain and sits in the public domain, so there is no serious technical moat protecting any one company. He points to Cisco as the cautionary tale, a real and important infrastructure company that still lost 92 percent of its market cap when the overbuild corrected, and argues something similar could hit the AI names even though they have genuine value. He puts numbers to it, sketching OpenAI's burn against its revenue, and reads the gap as the kind of thing that forces a reckoning. His history at Intel Labs informs the disappointment he voices at labs that started on an open-science charter and now argue for closed science, which he treats as a red flag rather than a strength. He closes on the outcomes he sees as plausible over the next six to nine months, including a Microsoft-Anthropic alignment or a Google-OpenAI acquisition scenario, touches on Yann LeCun's world-models direction as a reason the current architecture is not the end state, and advises a small-local-model backup plan given the odds that neither incumbent stays on top. 🔍 Topics covered: - Prof G's 1999 parallel and the two-player demand problem - Meta leasing out capacity it overbuilt - Why public-domain transformers leave the labs without a moat - The Cisco 92 percent drawdown as the cautionary tale - OpenAI burn math against its revenue - Open-science charters turning into closed-science arguments - A six-to-nine-month window and possible acquisition scenarios - Yann LeCun's world models and a small-local-model backup plan 💬 Do you think either OpenAI or Anthropic stays on top through the next couple of years? 🔔 Subscribe for no-hype tech analysis: https://youtube.com/@gojutechtalk 📺 Related: Why Dario Amodei's Argument Against Open Source AI Misses the Point https://youtu.be/JR22KE6kLMA 📺 Related: The Dangers of Over-Reliance on LLMs and AI https://youtu.be/U1Dhfij4Uy0 📺 Related: The Problem With Today's AI (In Simple Language) https://youtu.be/Cl7x2OhbPwU #AIBubble #ProfG #OpenAI #Anthropic #Cisco #Transformers #YannLeCun #GojuTechTalk