[Week 4 of 2026] Brands, Battles, Benchmarks

Welcome back to Price and Prejudice with a few musings from Week 4 of 2026.

McConaughey and the Machine

This WSJ article talks about how the actor Matthew McConaughey trademarked a few carefully chosen vibes and catchphrases of his. This is a good reminder that you can either get disrupted by technology, or you can file paperwork and try to charge rent for being you.

It is a nice illustration that “brand” is where the money is. AI makes imitation cheap, but it doesn’t make recognition cheap, so the rents accrue to people whose faces and voices already mean something to lots of people. If you are a generic actor, AI is a threat; if you are a distinctive actor, AI is a great opportunity. The deeper point of the story is about default ownership, and how it quietly flips in a world of cheap replication. In the analog era, you owned yourself automatically; in the AI era, you apparently need to register yourself like a logo or a sandwich.

Pokémon Benchmarks

The latest benchmark for artificial general intelligence isn't a complex math proof or a Turing test, but a rather humble Pokémon game. This is a nice reflection of the gradual progression from the "chatbot" era to the "agent" era, and Pokémon is the perfect crucible because it requires state persistence: the AI must remember that it bought a Potion in Pallet Town to use it hours later in Cerulean City. It turns out that this is a pretty good proxy for what people actually want from AI, which is not brilliance in short bursts but competence over time.

The thing people miss is that Pokémon is quietly game-theoretic in a very human way. You’re constantly modeling other agents – gym leaders, rivals, random Pokémon in the grass – and making tradeoffs under uncertainty with incomplete information. Do you grind now or push ahead? Do you optimize or explore? AI struggling with Pokémon isn’t failing at a game, but rather it’s learning that intelligence is mostly about anticipating incentives and muddling through anyway. And this has a nice parallel for financial markets, which are, after all, just very complicated Pokémon battles where nobody tells you the type chart.

Academic Frameworks

This note from Morgan Stanley is basically the Capital markets and investments syllabus I teach at Columbia, where it lays out the core questions (what efficiency even means, where it breaks, who’s paying the costs, and who’s getting paid) and gives you useful frameworks. If you want students (or investors) to stop arguing about whether markets are “efficient” and start asking how and when they aren’t, this note is a great starting point.

The academic instinct on display here is the search for unifying frameworks: behavioral, informational, analytical, technical – all clean buckets for a messy world. Yet, no single framework wins all the time because markets evolve and incentives mutate. Sometimes behavior dominates, sometimes flows, sometimes leverage, sometimes nobody knows what’s going on but everyone trades anyway. Which is why it’s actually kind of astonishing when a media article confidently explains why markets moved on a given day – because even with hindsight, that causal story is usually much harder than it looks.

Podcasts

• This episode of Odd Lots (October 2025) discusses the use of AI at a major trading firm, offering a rare glimpse into how generative AI tools may be integrated with human traders.