This edition explores the potential of locally sourced language models, emphasizing the DIY spirit and independence in AI development. It discusses the vibrant community of enthusiasts striving for performance parity with foundation models, and the implications of running local LLMs on personal hardware. The newsletter invites readers to join the Season 7 opener of General Seminar, where the focus will be on translating weak signals into immersive experiences using Design Fiction. It also highlights the intersection of AI...

This issue invites you into General Seminar S07/E01 (January 14, 10:00 UTC-8) to take the updated AI 2027 Futures Model as a provocation: not something to worship or dunk on, but fuel for critical thinking and speculative imagination. We’ll translate model claims into everyday evidence — design fiction artifacts you can “hold,” like future-archeologists’ finds from a near future where AI is as ordinary as breakfast and as ominous as a USB cable. Alongside the seminar invite, there’s...

This Week 52 issue leads with a story about audacious containers and worldbuilding: a young musician asks what it means to be a musician in the age of AI, and the reply arrives sideways via a design-fiction “album” that’s not pressed to vinyl at all, but built as a box of artifacts plus a period-evocative website that feels like it fell out of 1984. The throughline lands in a blunt claim: creativity doesn’t happen in stacks, services, distribution...

This issue is a love letter to indexes: catalogs, archives, taxonomies, card files, and all the pragmatic ways we make complexity navigable. It frames the obsession as partly existential (how do you know your own work without maps?) and partly practical (indexes reveal patterns, relationships, and connections). It name-checks Bowker and Star’s argument that classification is always situated, then points to the Near Future Laboratory archive as “available but buried” and contrasts old-school index.html directory listings with today’s...

A welcome-to-new-subscribers issue that sprawls (in a good way) into “reader mail” replies to last week’s specialists-vs-generalists essay—clarifying why design thinking optimizes defined problems while speculative prototyping reframes unstable situations so the *actual* future gets discussable before it becomes emergency mode. Julian unpacks the “wheels on luggage” distinction, argues that range changes what you notice and experience changes what you trust, and frames speculative prototyping as an organizational early-warning capability rather than a one-off workshop. Along the way:...

This issue uses two decades of “end of the university” discourse to frame a deeper problem: our institutions are structurally misaligned with the futures they claim to prepare us for. It argues that futuring is not a matter of better forecasts or more specialized credentials, but of pairing domain experts with generalists who can see across silos, notice misalignments, and prototype new ways of organizing human potential. Alongside this, it shares an immersive harbor-district future from Ars Electronica’s...