For much of this year (2014) I've been noodling the idea of running a seminar series to discuss some fun questions about collective intelligence, cognitive tools, and so on.

I'm not sure on the exact theme. I started out with fairly specific ideas in mind, but recently have thought to broaden it out to something like "The future of computation". It has the disadvantages that: (a) it's far too big a subject to address completely; and (b) it doesn't provide a whole lot of guidance to me (constraints are good!). But on the other hand it is a nice big tent for some fun exploration.

The following document contains a few rough notes toward such a syllabus. It's very telegraphic and disorganized at present. It's also strongly oriented toward fundamental questions which I have a particular research interest in.

Designing cognitive tools

This is a strangely under-explored area. Of course, pieces of it are explored by the HCI community, by the graphics community, by industrial designers, user interface designers, and so on. But the design of cognitive tools per se seems strangely under-explored. There is much of interest in a Google Scholar search – but less than one might think, given the topic.

Bret Victor's work, particularly Kill Math (my notes), Up and Down the Ladder of Abstraction, Media for Thinking the Unthinkable, Learnable Programming, Inventing on Principle, The Future of Programming, Seeing Spaces, and Magic Ink. Seymour Papert's work is a great complement, especially "Mindstorms".

Engelbart on Augmenting human intellect, Bush on As we may think, Nelson on Literary Machines. Licklider's best known work is on Man-Computer Symbiosis. It leaves me a bit cold, though; I should re-read (and maybe warm up to it), or find something else. Alan Kay's work. Maybe his early paper on the DynaBook? Some of the recent stuff coming out of VPRI might be great to look at. His history of Smalltalk is very stimulating.

Robert Ochshorn's work. Maybe Joel de Rosnay on The Macroscope. Maybe something on systems thinking. Hard to say what: Donella Meadows, perhaps, or Jay Forrester? In general, there is a big question here: how can we develop tools to let us see and navigate our entire culture? We're at the beginning of doing that, not the end. A related point: we know almost nothing about the cities we live. How can we develop tools that reveal our own cities to us? Yelp and Google Maps are a tiny, tiny start. There might be some interesting stuff in the digital humanities literature.

Something people don't understand is that user interface design is an art, at least as difficult as writing an outstanding book or creating a great movie or painting. Indeed, in a sense the latter can be viewed as special cases of the former. I'd like a reading that conveyed that sense. Better yet, maybe set a project for people to do. It'd also be great to have something that conveyed just how open-ended a problem user interface design is, and how constrained are the usual ways people think about it. Bret Victor's Magic Ink is a good start on this. Maybe Don Norman's "Things that Make Us Smart" or "The Design of Everyday Things"? Maybe Raskin on "The Humane Interface"?

Programming languages as user interfaces. The hundred year language. What else is good on this? Maybe Guy Steele's "How to Grow a Language"? Why did programming language design stagnate in the 1970s?

Designing new forms of collective behaviour

The network enables us to create new forms of collective behaviour. I believe that: (1) we understand almost nothing at all about the design of collective behaviour; and (2) it's possible to create processes and technologies which dramatically expand the range of possible collective behaviour.

Some examples of new modes of collective behaviour: Yanis Varoufakis's analysis/ of Valve's unusual management model. See also the Valve manual, linked from that post. Other companies with unusual management models include GitHub, Gore, Mondragon (a collective), and Semco.

Open space technology (Harrison Owen) and offshoots (unconferences, ignite, etc). The Matlab Programming Competition (see Ned Gulley's papers on it). Maybe "This is Burning Man". Kasparov versus the World. William Whyte on "The Organization Man". Liquid democracy. Other alternatives to democracy. The point is to start to break down some of people's very constrained points of view on group process.

Prices and markets as enablers of collective action: I, Pencil. Hayek on prices and The Use of Knowledge in Society. Adam Smith's original explanation of the invisible hand. Possibly Stinchcombe's "When Formality Works". Possibly Frank Pasquale's (forthcoming) The Black Box Society: The Secret Algorithms that Control Money and Information. Red Plenty (with enough computing, a command-and-control economy may work). How Bitcoin works. Mike Hearn's talks on contracts. Dominance Assurance Contracts as a way of providing public goods (this could utterly change the world, extending markets so they can provide public goods). What are the ultimate limits of financial instruments? Maybe Kevin Slavin's talk about bots? Maybe some of the ideas people have about virtual corporations? It might be nice to describe very decentralized protocols like Bittorrent as well.

The above separates out group process from financial instruments. But these two things are not really separate from one another. It'd be nice to have a frame to unite them.

Theoretical frameworks: Coase on "The Nature of the Firm". Benkler on "Coase's Penguin". Probably Eric Raymond as well (maybe "The Cathedral and the Bazaar"). What Benkler gets wrong (or doesn't see). Ostrom's wonderful book "Governing the Commons" as a framework for thinking about collective action. Schelling (Micromotives and Macrobehaviour) and Axelrod (The Evolution of Co-operation) and Harding (The Tragedy of the Commons) for contrast to Ostrom. Lessig on "Code as Law". It'd be good to look for some followups to Lessig, too. Another possibility: Herb Simon on "Administrative Behaviour" (or maybe one of his other works).

Other stuff

Questions: What comes after the web? How to allocate expert attention? (Connect this to the work on mechanism design, particularly the various theorems on optimal matching). What design principles should we use for cognitive tools?

Cory Doctorow on The Coming War on General-Purpose Computing, especially the second half. Vinge's singularity paper. Michael Goldhaber on The Attention Economy and the Net contains some stimulating ideas about attention as currency. Bostrom's new book may be worth looking at (I haven't, yet). Herb Simon's "The Sciences of the Artificial" is great framing for any discussion of the future of computing.

I've largely ignored AI above. That's a mistake. It actually looms over the whole discussion. At some point it'd be good to imagine various possible AI futures (as in Vinge's Long Now talk). Which future we end up in determines a lot about the above.