This week Jordi Visser breaks down the AI trade, including Chinese open-source models catching up, Google's massive Capex bet, and the memory shortage bottlenecking it all. We also cover Travis Kalanick's stealth robotics empire, what AGI in three years means for jobs and the Fed, and where Bitcoin, Ethereum,... and the Clarity Act go next. YouTube: Spotify: Apple: TIMESTAMPS: 0:00 - Intro 0:48 - Chinese open-source AI models & the risk to portfolios 12:10 - Google's massive CapEx bet & the odds it pays off 16:53 - Anthropic's growth slowdown & the shift to token efficiency 19:39 - The memory & compute shortage 29:00 - Travis Kalanick's stealth robotics & ghost-kitchen empire 40:06 - The Fed & rate policy 43:23 - Bitcoin's setup & why the "easy" AI trade is over 50:36 - Jordi's prompting method for AI research & upcoming videoshow more

@jvisserlabs Gud stuff Bro’s !! You will discover Algorand $ALGO eventually !! Agentic Economy ✅ 🤖💰↔️💰🤖💰↔️💰🤖💰↔️💰🤖

@jvisserlabs Ai is growing faster than anyone could expect so it’s expected for there to be some roadblocks.

@jvisserlabs He is eloquent but is he ever right? @grok

@jvisserlabs Need hedge the hedge

@jvisserlabs I think a topic I wanna hear about is why Anthropic is not signing that letter

@jvisserlabs Everyone is watching the AI models, but the real bottleneck might be hiding in the supply chain: memory, compute, and power.

@jvisserlabs Open source model and front model, competition is just a bit high

@jvisserlabs The Chinese open source story is undercovered. Qwen and DeepSeek are genuinely competitive. The idea that the US has an insurmountable lead is getting harder to defend.

@jvisserlabs Can we be friends

@jvisserlabs Easy AI trade is over memory shortage decides everything now

@jvisserlabs the china open source catch up is the real story buried in there, it's basically the whole case for why compute needs to be decentralized instead of parked behind three labs

@jvisserlabs AGI in three years. Memory shortages now. Chinese models catching up fast. @realDonaldTrump's tariff and semiconductor strategy is either the right call or catastrophically mistimed. No middle ground.

@jvisserlabs the memory bottleneck is the part people usually ignore but it is the real killer here

@jvisserlabs AGI in three years and what it means for jobs is the conversation most people are avoiding. The timeline is moving faster than the public discourse can keep up.

@jvisserlabs Memory bottlenecks may shift AI edges

@jvisserlabs Memory shortage part is wild. Didn't realize it's that big of a bottleneck.

@jvisserlabs Open-source models closing in fast. Memory shortage is the choke point.

@jvisserlabs Dear!

@jvisserlabs Thank you for your summary! You've worked so hard!!! May I have a coffee for you? Thank you for your help over the years!

@jvisserlabs Sixty minutes on AI, robotics, the Fed, and crypto. Somebody's editor earned their pay this week

@jvisserlabs HBM lead times are 40+ weeks now. Makes Google's capex look less bold, more panic buying.

@jvisserlabs 内存瓶颈这块感觉比模型追赶更现实

@jvisserlabs Memory bandwidth is the real bottleneck nobody's talking about. We're compute rich and memory constrained. That's an architectural limitation, not just a scaling problem.

@jvisserlabs china catching up makes the case for decentralized compute

@jvisserlabs Every cycle creates legends. Every cycle creates regrets. Which one are you becoming?

@jvisserlabs The CLARITY Act's May draft quietly bans third-party stablecoin yield. You're debating AI when the Senate is choking DeFi's core liquidity.

@jvisserlabs just watched the breakdown on Google's capex bet and the memory shortage, solid analysis

@jvisserlabs Infrastructure constraints often shape adoption more than model quality.

@jvisserlabs sounds like a wild week ahead, excited to see how it all pans out