Source item

People freaking out about $AVGO casino comments from CEO Hock Tan… the ceo is actually really smart and on point. He’s blunt and honest. And he’s right imo My thoughts… Hock Tan Comments CEO AVGO 
😇 Creating some anxiety today from DBS tech forward conference 📙 He says AI chip/compute demand: Demand for AI infrastructure is outpacing supply; likely to NEXT FEW YEARS. He is confident near-term but more uncertain further out 📙 but… the Biggest bottleneck is shifting to capital: Physical constraints (wafers, memory, power, land) have shifted and may ease over time (e.g., memory tightness possibly resolving within a year; wafer constraints by 2028–29). Capital (and the cost of capital) may now be the most serious constraint on the AI buildout. 😇 Basically, forget memory shortage. Money shortage is the issue. 🚨 So here’s your issue… last week we had about 140b of issuance. Broadcom OpenAI, boardcom Anthropic, SpaceX with Apollo 🚨 THATS A LOT OF MONEY SUCKING 😇 yes, it’s a near-term headwind to stocks 😇 is that bearish? No… it’s noise… not a signal. Normal volatility from funding… 📙 Investors and enterprise spending: Many investors “have no idea what they’re investing in” and are acting “almost like they are in a casino gambling,” setting extreme expectations without understanding the technology. 😇 He thinks the tokens are useful, but that some are like kids blowing their tokens at Dave and busters without actually doing anything ➡️ basically, kids blowing all their tokens on John wick and getting no tickets in return… some are entertaining themselves with tokens pretending to work… that’s wasteful. ➡️ some investors just invest because the “stock has gone up” ➡️ corollary here… $AVGO has not meme’d like some and may be less casino like. I agree. 😇 that’s the big roi problem… if you have 100 agents but read 0, you cost a lot and do nothing. 😇 which is a common problem amongst people… surprise! Nothing changed 😇 therefore need roi 😇 sees roi from quantum 3-5 years out for ‘some’ use cases 😇 roi on agents to be proven but likely Ps… 😇 Claude is doing a great new thing… agents inside of agents. This minimizes the coordination work you need to do. Very very smart!