← All episodes
Work, Measured

We Optimized Away the Human Part of Work | Meg McWilliams

Six in ten people told Meg McWilliams they would not have attended at all without one warm introduction beforehand. The friction was the point.

When Meg McWilliams ran user interviews on an early version of her product, which did nothing more than send warm introductions by text before an event, roughly six out of ten people told her they would not have turned up at all without it. Their reason was not logistics. It was that they usually walk into a room where nobody is waiting for them.

Who is on this episode

Meg McWilliams has spent her career building things that bring people together, starting at the LA Film Festival and Sundance, then moving into VR and immersive experiences, then building communities in gaming. She is now the founder of Mixies, an AI native platform for in real life communities, which she describes as wanting to be Discord for events and groups that meet in person.

What they covered

Meg's read on the market is that the pendulum has swung back hard. Go to market plans are now built around getting more dinners happening in a city, and community roles have become some of the most sought after hires at AI companies, because standing out online has become close to impossible and people will not answer an email from someone they have never met.

The more interesting half of the conversation is about how the losses happened. Nobody announced them. Contactless delivery, cashierless checkout, a door you never have to open: each one is defensible on its own and the aggregate is a life with no unplanned human contact in it.

"It's not like there's some big thing on a billboard that's like you will not see anyone for the next two weeks. It's like all these like micro decisions that get made, in the name of efficiency."

Meg McWilliams

Imran brings the operator's tension into the open: a 45 minute commute reads as two hours of lost output, and that calculation is now running constantly in the background for everyone. Meg's answer is not to deny the maths but to point out what the ROI calculation leaves out, which is the part that explains why you were doing any of it. Her practical defence is a protected weekly ritual, borrowed from a Kurt Vonnegut story about going to the shop to buy a single envelope rather than ordering a hundred, because the errand is the point.

"What's my fart around moment that I can go and like just be in the world?"

Meg McWilliams

She has never owned a printer for exactly this reason. It gives her a standing excuse to go to the library and talk to the librarian.

On AI she is deliberately unromantic. It is a tool, and the question is which direction you point it. She is genuinely glad that legal advice is being democratised, because six hundred dollars an hour to ask a basic question is a real tax on a founder. She is also honest about what that removes: at her first company she had a relationship with her corporate lawyer close enough that she cried on the phone to him. Her preference is for AI that reduces the manual work so there is more time with people, rather than AI that removes the person.

Key insights

  • One warm introduction before an event changed attendance for roughly six in ten of the people Meg interviewed. The barrier was not interest, it was the fear of walking into a room where nobody is expecting you.
  • Efficiency losses arrive as small individually sensible decisions, which is exactly why nobody catches them. There is no single moment to object to.
  • When every interaction at work exists to extract a specific piece of information, you have removed the conditions for trust, not just the small talk.
  • Put a protected, inefficient errand in your week on purpose. Meg's version is refusing to own a printer so the library trip stays in her calendar.
  • Point AI at the manual work, not at the relationship. Her test is whether the tool gives you more time with humans or removes a human from the loop.

Watch the full conversation on YouTube, or listen to the audio version on Apple Podcasts.

Find out

Find out what you actually have.

Thirty days, one named engineer, and a readout of what your team and your agents actually produced. $2,500, credited in full against your first year.

Got it · we will reply within one business day
That didn't go through · check the address and try again
Built by operators, not HR software peopleSOC 2 Type IINever trains on your data