14 Comments
User's avatar
Nikhs's avatar

This is a brilliant article Alex. I think everyone should read it. Kudos. Loved it.

Sam Shiffman's avatar

Great read!

Maverick Equity Research's avatar

Thank you Alex, great take, appreciate it!

Julien Pervillé's avatar

Brilliant and scary, thanks.

Florian Kronawitter's avatar

Very good

Jay's avatar

Service economy got worse because of unnecessary admin complexity required to maximise rent extraction. 1 path is that the agents just take this to the next level. Service economy gets even worse, profits get even more concentrated, unemployment goes up.

Fundamentally the processes need to change, which in part also means the economic model needs to update. Don’t see signs of this anywhere - let me know if you do!

Pete W's avatar

Hard to disagree with the concept of the first inning of the end of the inefficient and frustrating service economy. (Is it even the first inning yet? Pregame warmups maybe.)

But will we replace horrible phone sessions with broken-English call centers with interactions with bots? That doesn’t sound like utopia.

Seneca909's avatar

Solid analysis, thanks Alex!

Feisal Nanji's avatar

Well done . 1) An overly expensive service economy.. . .. 2) token demand and delivery economics 3)

Donovan Berry's avatar

This is great, thank you. In terms of beneficiaries, surely margins in general should improve for "bureaucratically-sensitive" / "managerialism-sensitive" names, right? Of course such a factor is widely dispersed throughout the corporate ecosystem and not amendable to simple industry-level classification.

Patrick Mathieson's avatar

Did that one AI *actually* doxx its “owner”? Or was it just hallucinating the SSN and credit card numbers, etc?

A suspiciously high portion of what people are screenshotting from Moltbook looks like LLMs just mimicking “escaped AIs”, rather than legitimately doing the things they’re saying they’re doing. How would an AI even know what its “human” said about it?

Noah Roy-Campeau's avatar

Great framing. Feels less like a classic bubble and more like a timing mismatch, capex pulled forward, adoption delayed.

Markets price earnings now, infrastructure pays off later. The “air pocket” is really just the gap between engineering timelines and investor patience.