More or Less · Topics · Episode #174

Sam: To Understand the Agentic Future, Read Ari Emanuel's Book on Hollywood Agents

Clip from the full recording, 24:35–29:34 · download mp4 · watch at this point on YouTube

Sam argues AI agents will work like Hollywood agents: a few trusted relationships doing whatever you ask. Parker counters that the moat goes to whoever sees the work that needs doing first.

Sam says the best way to understand the agentic world is to read Ari Emanuel’s book about actual agenting. The agentic future will work like real agents: you have a few trusted relationships, and when you tell one what you need done, he does it. That is a break from the software past of a core competency in one vertical. The risk, Sam notes, is that Hollywood agencies are a terrible business where you get fired and replaced all the time because it’s trust-based with no competency moat. He sees startups with barely a strategy who go get customers first and just build whatever they need.

Parker agrees the question is where the moats are, and offers an unusual view: the most important moat will be visibility into the fact that there is work to be done. There’s a huge gap between AI capabilities and uptake because most people don’t know what they can hand off, so the winners are problem identifiers who bring the solution before the customer asks.

He argues Rippling is an unusually rich source of such triggers because employee lifecycle events drive so much G&A work. His example: a terminated employee got a signing bonus with a clawback, so Rippling can notice and offer to recover it.

Key points

  • Sam: to understand the agentic future, read Ari Emanuel's book on Hollywood agenting.
  • Sam: agents will work like human agents, a few trusted relationships doing whatever you ask, not narrow vertical competence.
  • Sam: Hollywood agencies are a terrible business because the trust-based work has no competency moat; you get fired and replaced all the time.
  • Sam: new startups barely have a strategy; they get customers first and then build whatever those customers need.
  • Parker: the most important moat will be visibility into the work a business needs done.
  • Parker: most people don't know what they can hand off to AI, so winners will identify problems before the customer does.
  • Parker: employee lifecycle events, like a terminated hire who owes back a signing bonus, are rich triggers for Rippling.

Where they landed

SamAgentic business will look like Hollywood agenting: trust-based, many competitors, and lanes of competence blur.
ParkerThe moat is upstream visibility into problems before customers ask.
JessicaProbes whether this model means agents are probing and gathering information on people.

Quotes

“if you want to understand the agentic future, the best thing to do is read Ari Emanuel's book about actual agenting, right, in Hollywood.”— Sam
“You get fired all the time, you get replaced all the time. It's a trust-based thing. There's no competency.”— Sam
“They barely have a strategy. What they do is they go get customers first.”— Sam
“the most important moat is gonna be which companies can have visibility into the fact that there is like work to be done for a business.”— Parker
“if you're waiting for the customer to bring you something, it's too late. Someone else has already solved it.”— Parker

Suggested tweets

Sam Lessin's hobby horse: "If you want to understand the agentic future, the best thing to do is read Ari Emanuel's book about actual agenting, right, in Hollywood."

Tweet this →

Sam on the startups he's seeing in the AI era: "They barely have a strategy. What they do is they go get customers first."

Tweet this →

Parker Conrad: "If you're waiting for the customer to bring you something, it's too late. Someone else has already solved it."

Tweet this →

All topics · Full episode #174 + transcript · Subscribe on YouTube · Spotify