Adam Swick

🚨 PSA 🚨 I’ve received reports of fake telegram messages impersonating me and fake Zoom calls using someone who looks like me. I will never ask you to upgrade Zoom, install software, or click links. If something feels off, hang up and text/email me directly. Most all of you have multiple ways of getting in touch with me, so 2FA. Stay safe.

Three companies quietly control one of the biggest bottlenecks in AI. Not chips. Not models. Memory. This will matter more for geopolitics than most people realize.

Memory used to behave like a commodity. It now behaves like infrastructure. Few suppliers. Massive capex. Multi-year timelines. Strategic choke points. Treat it like oil or power

20,000,000 Bitcoin have now been mined. Only 1,000,000 BTC left to be created. Ever. But because of halvings, that last 5% will take ~114 years to issue. ~94% of all BTC already exists. The remaining supply will trickle out slower and slower until ~2140. Perfectly predictable monetary policy. No committee. No bailouts. Just math.

People keep asking when the AI memory cycle peaks. Wrong question. This is not 2018. Supply is tighter. Scaling is harder. Demand is non-linear. Cycles driven by physics last longer than cycles driven by optimism.

For decades, memory was deflationary. Smaller nodes. Cheaper bits. Rinse, repeat. That era is over. Memory pricing is now driven by physics, capex discipline, and long lead times. That changes everything about AI economics.

Read a lot more great detail here: newsletter.semianalysis.com/p/are-ai-datac….

Everyone is blaming AI datacenters for rising electricity bills. But the data suggests something more nuanced: the biggest driver may actually be electricity market design, not AI demand. Two regions illustrate this clearly: PJM vs ERCOT. PJM operates the grid across 13 eastern US states. It uses a system called a capacity market, where power plants are paid to be available in the future during peak demand events. The price for that capacity is determined in auctions that rely heavily on forecasted demand for electricity years in advance. When PJM modeled future load growth from AI datacenters and hyperscalers, those forecasts shifted the auction curve dramatically. Capacity prices jumped 9.3×, which is now flowing through to consumer electricity bills. The result: households in the region could see ~$25–30/month higher power costs compared with a few years ago. Now look at Texas (ERCOT). ERCOT does not run a capacity market. Instead it uses real-time scarcity pricing. Electricity prices rise only when supply actually gets tight, rather than being determined by forecasts of future demand. Texas is also experiencing enormous AI datacenter expansion. Hyperscalers, AI labs, and GPU cloud providers are all building major infrastructure there. But electricity prices in ERCOT have increased only modestly, roughly in line with normal market movements. Same AI boom. Completely different price outcomes. The difference is that PJM’s system allows forecast models of future demand to directly influence prices, while ERCOT’s system relies more on actual supply and demand signals in real time. This doesn’t mean datacenters have no impact on power systems. They absolutely do. But the evidence suggests that how electricity markets translate demand into prices matters just as much as the demand itself. As AI infrastructure scales globally, the competitive advantage for datacenter regions may not just be cheap electricity or available land. It may be how their power markets are structured. In other words, the bottleneck for AI infrastructure might not be electricity. It might be energy policy.

Two assumptions we all made: OpenAI would dominate models NVIDIA would define the entire stack Neither looks guaranteed anymore. Optionality is back.

“Green data center” is marketing until regulators define it. Bring your own power, emissions plan, and community upside or expect rejections to accelerate.

NVIDIA can ship every GPU it wants. It won’t matter if power and land aren’t there. The AI race is now a grid-engineering problem.

The neocloud pivot is obvious in hindsight: GPUs → managed training → managed inference → production SLAs Compute without operational context is a race to zero.

Our Line Up for Today | Feb 16th, 2026 starting at 11:00 AM EST 🎙 11:00 AM: Sanjna M ( @sanjna_connect ) 🎙 11:00 AM: Natalie Brunell ( @natbrunell ) 🎙 12:15 PM: Adam Swick ( @swicktalk ) Make sure to set your alerts!

AI wants to scale at software speed. Gas turbines, alloys, transformers, and skilled labor do not. The next bottleneck is not GPUs. It is engines and steel. Bitcoin mining hit similar walls in 2021. AI is hitting them at 10x scale. Watch OEMs quietly become kingmakers.

AI firms are willingly paying more for onsite generation. Why? Because waiting costs more than emitting. Bitcoin miners learned this flaring gas and curtailing load. Imperfect power now beats perfect power later. When marginal time equals marginal billions, the debate ends. This isn’t ideological. It’s arithmetic.

If you could buy every empty seat on a plane, last minute, cheap, and give it back when someone pays more, airlines would love you. @swicktalk says that’s what Bitcoin miners are to power grids⚡️ New podcast live @ 12PM EST 👇

Thanks for having me on, @bramk! Bitcoin mining is the first large-scale load you can curtail in seconds, which makes grids easier to balance, unlocks stranded energy, and can lower the cost of building for peak demand. We also get into why AI data centers and miners look similar but behave totally differently today (uptime, cooling, capex), and how that converges over time.

xAI made it obvious. Everyone else is now copying the same move. Mobile turbines. Behind-the-meter gas. Rented engines. The grid is no longer the product. It is the backup plan. Time-to-power beats $/kWh by orders of magnitude. Bitcoin miners built this muscle in 2018. AI just scaled the stakes to billions.

Interesting contrast right now. Crypto finally getting regulatory clarity in DC. AI data centers fighting state and local battles over power and water. One sector is graduating. The other is discovering NIMBY reality. Miners who survived zoning fights, grid politics, and curtailment rules have a real advantage in AI infra. Decentralized compute isn’t new. Bitcoin already shipped it. AI should steal more ideas.

People keep asking if AI is a bubble. Wrong question. The bubble is paper megawatts and interconnection queues. If your AI roadmap depends on a grid upgrade in 2029, you don’t have an AI business. You have a slide. Value lives where electrons are already flowing.