Matt and Jack are co-founders of HeySquibb, an AI research lab modeling human psychology to build digital twins that run your social media. After Matt built an AI therapy clone in 2023 and saw the impact a digital representation of someone could have, the team came together around a bigger thesis: that within a year or two, having a digital twin will be as ordinary as having an email address. They're bootstrapped, fully self-funded, and recently moved six teammates into a four-bedroom apartment in New York City for three months to live, sleep, and ship together.

In this conversation, Matt and Jack walk us through how their digital twins reach 76% accuracy on you, why ”sounding human” is the hardest problem in AI right now, the four-product roadmap (Persona, Warehouse, Telescope, and the Omega Suite), why they think AI agents will free us to be more human (not less), the ethics of agent-driven social engagement, and what it actually felt like to bootstrap with the bear in their own cave.

Find HeySquibb: https://heysquibb.com

On X: @heysquibb

Builders & Doers is produced by Horizon Search.

0:00 Intro

0:47 What HeySquibb is — digital twins to run your social

1:34 Matt's 2023 AI therapy clone and the origin story

2:14 Can it be a better version of you? 76% accuracy today

2:47 What if the data you've put online was already AI?

4:00 How long it takes to train your twin

4:48 Templates — sound like Mark Twain or Mark Cuban

6:14 Why X first, LinkedIn second

7:03 A world where bots talk to bots — and what that unlocks

7:16 The vision: get offline by being more online

8:13 The dark mirror — being intentional with what AI reflects back

11:08 The sales prospecting use case

13:01 Is using a digital twin ”cheating”? The calculator analogy

17:59 How the agent actually learns — human-in-the-loop reinforcement

19:53 Is AI a net positive for humanity?

22:33 The ”in or out” choice — and why dioramas always crash

23:07 Six guys, four bedrooms, three months in NYC

26:14 Bootstrapped from day one — putting your own skin in the game

28:50 The moat against the incumbents

32:08 The 12-month roadmap: Persona, Warehouse, Telescope, Omega Suite

32:50 API risk and staying compliant on X and LinkedIn

34:27 The hardest technical problem — keeping agents from going off the rails

36:21 Humor and context — the edge cases you can't fully anticipate

42:00 Security, privacy, and anonymizing user data

43:54 Will AI sterilize marketing for the human race?

47:00 Digital twins as snapshots in time — the expat analogy

48:23 What if two digital twins have a ”child”?

52:24 How bootstrapping actually works when the month is looking low

54:50 The bear in your cave — the strongest motivator

56:30 Where to find Matt, Jack, and HeySquibb