More or Less · Topics · Episode #163
You Can't Financially Price AI -- It Trades Like Moderna, Not Like Electricity
Clip from the full recording, 33:57–39:33 · download mp4 · watch at this point on YouTube
Sam argues the industry's favorite comparisons (electricity, the early internet) don't work because AI has no moats or lock-in, and reaches for a better analogy: Moderna, a great trading story that's a terrible long-term hold.
Picking up on Jess’s observation that everyone in finance is now obsessed with whether Anthropic’s ARR is slowing, Sam admits nobody actually knows – and that not knowing is itself the structural problem, because a company that can slow down and then reaccelerate is nearly impossible for investors to price. He rejects the electricity comparison that AI executives (like CFO Sarah Fryer) reach for to justify scale: electricity turned out to be a mediocre long-term investment even though it’s ubiquitous, because it’s a regulated utility with real barriers to entry and hard-to-transport infrastructure. AI is the opposite – “globally liquid with no borders” and easy to switch between providers – so Sam expects a race to the lowest-cost provider with limited durable profit for any single winner.
His preferred analogy is Moderna: it’s had incredible news-driven pops (like the day it jumped from roughly $50 to $150-175 on a skin cancer vaccine announcement) but is fundamentally “an IP company, the same thing as publishing” with no durable moat, so the stock is a great short-term trade on narrative and a bad long-term hold. He tells the story of a colleague, Yoni, who put his life savings into Moderna on conviction but didn’t lever the trade – meaning he left what could have been generational wealth on the table by holding rather than trading the narrative swings. Sam extends the framework to SpaceX itself, calling it “a great meme” that trades on story rather than cash flow, and predicts the real money in AI will be made by people trading the narrative, not by those who buy and hold financial assets for their cash flows.
Key points
- Sam: nobody knows if Anthropic's ARR is slowing, and that unpredictability -- accelerate, decelerate, reaccelerate -- makes these companies nearly impossible to price.
- Sam rejects the electricity analogy AI executives use: electricity was a poor long-term investment despite ubiquity, because it's a regulated utility with real moats and transport friction that AI lacks.
- Sam's AI thesis: it's "globally liquid with no borders" and easy to swap providers, meaning no durable pricing power for any single winner.
- Sam's preferred analogy: Moderna -- huge narrative-driven stock pops (a skin cancer vaccine announcement sent it from roughly $50 to $150-175 in a day) but no durable moat, since it's essentially an IP/publishing business.
- A colleague, Yoni, held Moderna on conviction without leveraging the trade and missed out on what could have been generational wealth -- Sam's point being you should trade these narratives, not just hold them.
- Sam calls SpaceX itself "a great meme" -- an asset that trades on story and headline risk rather than financial fundamentals.
- Sam: the real money in this AI cycle will be made trading narratives and news events, not from owning assets for their cash flows.
Where they landed
Quotes
“The electricity comparison sucks for them... it's a regulated utility with regulatory capture. It's hard to transport. AI is globally liquid with no borders.”— Sam
“Moderna is like an IP company. It's the same thing as publishing... The innovations might be amazing, but unlike Lily, they don't have any angle other than publishing, which is not a good business.”— Sam
“SpaceX is a meme. It's a great meme... it has every ability to pop out headlines that people are gonna go nuts for.”— Sam
Suggested tweets
“The electricity comparison sucks for them.” Sam Lessin explains why AI companies can't be priced like a regulated utility.
Tweet this →Moderna jumped from $50 to $175 on a skin cancer vaccine headline. Sam Lessin: great trade, terrible long-term hold -- and that's basically the AI thesis too.
Tweet this →“SpaceX is a meme. It's a great meme.” Sam Lessin on why the smartest AI-era money is made trading narratives, not owning assets.
Tweet this →All topics · Full episode #163 + transcript · Subscribe on YouTube · Spotify