The One-Person Orchestra

Jazz pianist Greg Spero runs a studio and a record label in Los Angeles. Most of the operational work is now handled by a fleet of AI agents running in parallel. His day goes like this: give the fleet a job, put the phone down, and play until it calls him back. This evening takes that loop apart and hands it to the room.

While the agents work, the piano plays

“I give the fleet a job, put the phone down, and play until it calls me back.”

That is how Greg Spero describes his day. He is a pianist who runs a recording studio and a label in Los Angeles, tours, and keeps releasing music. Most of that work now happens without his hands on it. What he does is issue instructions, judge what comes back, and play piano in the gaps.

The first half of the evening is technical. Greg will walk through how his system actually works, then the participants will install it on their own machine and configure it, which is the real starting point for anyone wanting to work this way. From there the group builds something together: a small application for co-creating music. By the end of the session it is deployed and live.

The second half is a solo piano concert, with that application in the audience’s hands. Scan a QR code from your seat and you are inside the performance. The set is improvised, the AI tools are part of the instrument, and what the room sends back shapes where the music goes.

Leave with an installation, not a demo

Nothing here is a slide-deck demo. Participants install the same environment on their own laptops and open their first session in the room. Greg’s parallel session setup, the way he phrases instructions, and how he decides what to keep all come home with you as a printed field manual with a full command reference, plus a starter kit. A 30-day path is provided for continuing afterwards.

Everyone chooses what gets built and performed

There is no prepared brief. The room discusses what is worth building and condenses it into a single sentence, and that sentence becomes the first instruction. Whatever comes out of this evening exists only because of who is in the room.

You can come and just enjoy the music

If you want to build, bring a laptop or a phone. One screen between two people works fine, so coming empty-handed and looking over a neighbour’s shoulder is also an option. Or skip the building entirely: Session 2 stands on its own as a concert, and coming purely for the music is a perfectly good reason to be there.

Who this is for

• Anyone who wants to put AI agents to work but is unsure where to start
• People running a business, a practice, or a label on their own or with a small team
• Creatives, engineers, designers, and producers
• Artists, musicians, and researchers interested in how AI and creative practice meet
• Anyone curious about new formats for live performance
• Anyone who likes jazz and contemporary piano
  

Timeline

Participants are welcome to join Session 1 and 2 from 16:00 or simply join from Session 2 from 18:30.
Timings below are a guide and may change closer to the day.

SESSION 1 | UNDERSTAND, INSTALL, BUILD

SESSION 2 | LIVE & INTERACTIVE