Lab Night went nomadic: David hosted from Bucharest, and explicitly licensed the group to continue without him whenever needed. A collective with a single point of failure is not much of a collective.
The framing went one level up. If AI keeps improving, where is the next bottleneck? Not chips, not energy: the ability of human organizations to absorb compound acceleration. Civilizations can retreat from breakthroughs; nuclear energy is the precedent. Small groups that practice absorbing new capability quickly, and show others how, are microscopic design choices with macroscopic consequences. That is the civilizational case for what we do on Thursday nights.
The most memorable artifact was the Collective describing itself. In twenty minutes, idea to finished video, an AI composed script, music, voice, and animation, and articulated us better than we had: “small groups, sharp tools, fast cycles… The lab is the work. Each lab night, members ship something, show it for five minutes, take questions, steal what works, pass forward what they learn.”
A consequence David drew: the CV is dying. You knock on doors with a repository of what you built, or a pull request, “and then you are irresistible.” Products attract users before pitch decks attract investors.
We welcomed a new member from Florence, a neurotechnology founder entering pre-incubation, who captured the moment precisely: execution used to be everything; now that execution is increasingly available, the idea has a different hierarchy, and sharing ideas matters more, not less. When he apologized for talking about his project, David interrupted: “It is not a bother. It is the reason we are together.”