2005-06-10 reboot day 1 doc searls: information is what informs/forms/changes us [the different that makes a difference] we are all authors of each other authority is what we give certain people to shape and form what we are # ben cerveny: play is what lets us sculpt a surface on the environment. it lets us figure out our special moves, our vocabulary. cf, animals at play. we have the ability to model abstract systems. game design is providing, for culture, the way of understanding dynamic system models. "on board memory" a way of storing metaphors about the world. chess is a huge metaphor store. it's a metaphor for feudalism. then people internalise the model, and chess is used as a metaphor source. [this is the second time metaphors have come up as a basis vector today. the first was doc searls.] once you have enough entities and relationships, you can start having feelings about the game, make forecasts and so on. it becomes a language, the symbolic world. there are emotions, non-quantitative things. tarot: the game falls, and what remains is a system for expressing. composition. in some games - driving games, say - you start off by selecting attributes. you decide that you're going to pursue a certain rulespace from a certain angle by setting variables. then that becomes actualised, it becomes performative. conscious of all actions having consequences. state of composition (the surface) <-> instance of play (experience) state machine. there is an internal thing of the game. we learn how to model the state machine, the artificial intelligence of the game. games, when people play them for a long time, abandon the metaphor. you just manipulate the mechanics itself. the accretion of data and metadata means that we're accumulating new entities with which you can play. [this is all very much game as operating on an inscription. very literate. is there an oral equivalent to gaming? or is gaming, by default, trying to find the internal weather system of something-which-is-separate-and-inscribed? is gaming a kind of conversation, an alternative method to science?] will wright's take on simulation: it's not a game, it's a toy. games have end states, qualitative rules. toys are meditative things you learn in, in a state of play. [maybe it goes beyond literature vs orality. it's inscribing something which is unreadable.] # jimmy wales: english wikipedia: >50% of the edits are done by 0.7% of the users (524 people) the most active 2% (1474 people) have done 73.4% of the edits the rules for how to the site works is not done in the code, but done in freeform by the community. there's a "Votes For Deletion" page, but it's not really votes because it's not enforced by the software. # back to the nature: a crash course in biomimetic IT mikkel holm s¿rensen background: - mikkel is a philosopher - phd in biomimetic and ambient intelligence (AmbI) from the IT university of copenhagen - colleague of reboot founder in Antics Ltd bigmother.dk blog/site on surveillance (empathetic aspects of tracking technologies) --> http://www.itu.dk/people/megel/ papers: http://www.itu.dk/people/megel/texts.html the darwinist turn of IT: - classic mechanism: needs are analysed; functionality is created in a top-down way (watchmaker) - darwinist: IT must adapt; functionality emerges from self-organising process (gardener metaphor) [I was saying earlier that the problem with evnt is that you lose meaning by forcing events into datastructures. the way to get around this is to keep the communities really small, and let people attach their own meaning to the symbols] mechanism vs complexity: mechanistic - almost decomposable (herbert simon): functions correlate roughly with structural modules - acontextual (closed systems) - atemporal ('tempora-phobic') - eg, technology, esp hardware, gets worse over time [but, i think: technology isn't standalone, it's part of human endeavour. and in that context, technology does get better over time. data is the lifeforce; the hardware is the body cells that are completely replaced every 7 years, well 18 months] complex - many-to-many functional organisation (eg crucial inter-modular dynamics) - feed off the environment - when they get non-dynamic, that's because they're dead - circular causality and global constraints - formation of a bird flock controls how the birds fit into it - but the birds themselves decide. circular - intrinsically historic: 'tinkered' and dynamic (open systems) biomimetics: - 'bio': living systems are paradigmatically adaptive - 'mimetic's: focus on organisation of adaptic dynamics (design heuristics, not solutions) - includes Evolutionary Computing, Swarm Intelligence, Evolvable Hardware biomimetic engineering: mimics natural structures; design from without biomimetic IT design: mimics self-organising adaptice processes; design from within why biomimetic IT design? - IT getting more pervasive and complex - labour intensive - difficult to diagnose/categorise failures - difficult to design for fuzzy needs - natural adaptive dynamics - yields robustness, fault tolerance, flexibility - explores new domains in design space AmbI ecologies - biomimetic Amb: assitance from multiple self-organised devices - self-organised bottom up: devices seek to meet internal norms through interaction with users - what if we made the same corner turn as we did with copernicus? make these agents orbit getting stuff right, rather than just doing something and us fitting into it - feedback from users determines fitness - implicity: frequency and duration of use - explicit: evaluation, tipping/punishment - successful services proliferate - functional ideas: robustness, resilience, pliancy, assistance on demand DELCA ghosts - disembodied location specific conversation agents - http://delca.itu.dk - a parallel world of ghosts competing for popularity - disembodied but location bound - motivated and focused: haunt because of unfinished business in the world - have character (easily recognised attributes) - appear unexpected or if summoned - supernatural, not unnatural - the more useful they are, the more they're allowed to appear give the ghosts a kind of maslow heirarchy of needs: - lower: power, memory, physical situation - higher: information access (databases, logs) - higher still: enhanced conversation skills - near the top: cloning/mobility ultimate ambitions: - ghost markup language - merging feedback and micro-payment: ecological economy ~= reputation economy biomimetics, the road ahead: challenges - new IT architectures - new dynamic materials - integration of control with structure. bridging hardware-software dualism - methods to guide adaptive processes without ruining them - new conceptualisation of design hurdles - technological autonomy vs human control: open ended systems and functional norms - ethical and legal issues of non-linearity - psychological issues: uneasiness with a probabilistic and complex technology - commercial issues # the long tail of fashion www.ullamaaria.typepad.com/hobbyprincess (I hope this talk is online) genmon_ is me, below genmon_: i like the idea of rfids in clothes [18:17] max__: angora is in east africa, dude [18:17] guan: ChrisDodo: it's still annoying that it takes 30 days to ship anything from china [18:17] genmon_: so you could stand on the tube, zap the rfids, and find out what underwear everyone was wearing [18:17] ChrisDodo: genmon_: you like the idea of rfid in anything [18:17] max__: s/east/west/, oh fuck it [18:17] tfs left the chat room. [18:17] guan: thomas told me that's part of the reason why we don't have reboot bags this time [18:17] dweinberger: rfid's in form of bangles: A million dollar idea AND a fashion statement! [18:17] ChrisDodo: who needs china, cheaper to produce locally for small volumes [18:18] max__: yeah, who needs china? [18:18] Aaaaandy: exactly chris - specialisation [18:18] ChrisDodo: ship the parts in volume, personalise locally [18:18] Aaaaandy: max: the chinese? [18:18] sturob: taiwan? [18:18] genmon_: i don't think IDs for clothes is a good idea [18:18] Aaaaandy: like cafepress [18:18] genmon_: it discriminates against people who combine [18:18] genmon_: better to have an open and concise description language [18:18] max__: Aaaaandy: i think they'd be happier in the west [18:18] sturob: taiwanese politicians anyway [18:19] genmon_: with an #include statement # engelbart demo addressing: markers at any point in any file, can point at a character and enter up to a 3 character name. then later, at any time when you'd have to point at a character and run a command, you can hold down the button and type in the name instead. [that's a your-own addressing thing] there's something key about the renumbering in a list too. because the hierarchic list is always numbered in the same order in the same way, if you know your own position you also know exactly the address which is "back one" or "up two." we currently have 2 kinds of addressing: global-named, and local-point. engelbart's NLS has different forms of addressing: - locally, you don't need to point. you can address things that you can't see, because everything is given a name relative to your current position - globally, filenames, with an anchor going into them if you know it (by position, which is automatic, or named) - locally by name - locally by pointing - globally, by having a hyperlink in the demo: information retrieval [heh, and this is exactly the addressing thing i noted above] everything is information retrieval, every time you change a mode this is on-screen. [my comments in brackets] 'hot' retrieval -- known destination direct -- explicit specification jump to identify jump to name jump link indirect -- implicit specification [information stored by the system about where you want to go] jump ahead/return [memory about what you've been doing recently] jumps referring to structure content analysis [the search filters over statements they had? not sure] 'cold' retrieval -- unknown destination direct -- hierarchy -- categorization [where you don't know what you want but can describe it] for example, what is the directive to get roman numerals for page numbers? see (nls,direc,1:SGDB) indirect -- keyworks -- associative reordering see (etc) [one of the things i like most about this demo is that things can go "off the screen." the screen is of a certain size, and it's composed. if files are too long, they run off. you can select a bit and say "start showing here," but there are viewing panes. most of the screen is fixed: a video, feedback area showing you what you're typing.]