« Technology, innovation and risk |
Main
| Bruce Sterling lecture »
April 15, 2006
Apologies, injuries and allegories
I apologise for the problems with duplicated entries coming up on this blog. I'm having trouble with my server at the moment, and while I've successfully deleted the redundant entries, I can't rebuild my site to get rid of them whilst these internal server errors keep appearing. I'm on it, and hope to correct the problem later today or tomorrow but I am somewhat in the hands of my host... may be it is time to find a new one....
I'm also suffering from absolutely hideous neck and upper back pain and can't get to a masseur till Tuesday. So I'll be sleeping with a hot water bottle for the next few nights. Joy.
But on the bright side, I have been engaging with some interesting stories and seeking meaning in gaming and science fiction. I attended the Conjure National Science Fiction Convention today in Brisbane, partly to hear Cory Doctorow's second presentation on his tour downunder, but also to hear some new theories of game development and revisioning of the real versus the imagined self in online gaming. Cory's presentation was fantastic and varied dramatically from his Ideas Festival session, even though the subject matter was still associated with Digital Rights Management. In this new presentation, which is being groomed for delivery at the Ray Kurzweil Singularity Summit at Stanford next month, Doctorow considered the singularity direction that was being pursued by agencies of content production, and considered how the humans and corporations were going to respond to a period when technological capability will outstrip human intervention - essentially when the age when technologies no longer need the humans in the mix to solve problems or create content. Doctorow argues that this singularity event is not, in and of itself, feasible for the same reasons that digital rights management is no longer feasible. Content protection under digital rights management came about due to a culture of scarcity of content. At a crucial period in history, policy and content production came together to create copyright and rights management architectures to protect an industry of content production. Doctorow argues that industry is now mature, and the rights management system is now due to be retired. Indeed, he argues that attempts to accommodate rights management in a system that is technologically increasingly complex is actually killing innovation. He is adamant that complex systems require diversity and mutation in order to be sustainable, and he points out that even if that moment of the singularity event arrives, machines just don't do diversity well. He believes that the machines will still require the organic diversity of human interaction as a means of providing the content production sustainability, so Creative Commons and a philanthropic economy (as opposed to the more traditional concept of a gift economy - see last post) is likely to emerge. Some interesting theories there.
Later in the day today was a session on 'Game Theory and Emergent Behaviour/Divergent Stories', which I have to say I was a little disappointed in, because the discussion was more about which games were making money, not about which games had merit as arbitrators of divergent story telling. The speakers also really ignored the branch of mathematics which excited me in the session title: Game Theory as a strategic measure, designed to maximise returns for stakeholders. If the session had been more embedded in theory and less in financials I think I'd have enjoyed it more. Better (and more dynamic) was the session after lunch on 'Better than the Real Thing', which traversed the territory of the blurring line between reality and gaming. Questions about citizenship, participation, scenario mapping, corporate control and the exploration of a 3rd place/3rd way all peppered the session, and while insufficient solutions were proposed, the questions were much more pithy.
Tomorrow I'll be attending Bruce Sterling's lecture, which I hope will be a nice mix of his work on the hacker culture and his vision of science fiction as the semantic sweetspot for redefining culture. Stay tuned for a blog summary tomorrow.....
Posted by jj at April 15, 2006 4:44 PM
Comments. Please PREVIEW your comment before submitting. This helps the server to save your entry.
Machines can definitely "do diversity". The real dangers of the intelligence singularity are (i) pseudo-super-intelligent machines with human-indifferent goals (ii) a cul-de-sac scientific culture which thinks thought is computation. (Actually, there are surely other "real problems", but if we don't solve those two, they may not get to arise.)
Eliezer Yudkowsky's Singularity Institute is at least trying to solve the first problem, with its notion of "Friendly AI"; although his current notion of a Friendly goal seems to be, "figure out what humans currently want, figure out what humans would want to want if they really understood themselves, figure out what these humans-at-their-best would want you-the-AI to be and to do" - (this is the AI's thought process, remember) - "and then do that".
And this is not enough because of the second problem, the cultural and intellectual problem. We don't understand consciousness (and hence we do not really understand thought, memory, understanding, or any other "cognitive" process, insofar as it involves subjectivity); and we cannot understand it without getting outside the whole mentality on which natural science is currently based, which assumes a-priori what the philosopher Gustav Bergmann called a "truncated" ontology, an ontology in which everything "mind-like" is excluded by construction; and AIs as presently conceived have no capacity to do this, so the first step above - "figure out what humans currently want" - is already doomed to produce an ontologically deficient account of human nature. I cannot say exactly what hazards might arise downstream as a result, but they could be as bad as anything coming from a human-indifferent superintelligence.
I've written a little about the ontological issues in the comments here, but to really do justice to them will require serious philosophy, of the sort that directly faces the basic ontological ignorance that human beings have about the nature of reality, while making the most of what conscious reflection can actually tell us.
Posted by: mitchell porter at April 16, 2006 11:45 AM
Post a comment
Please PREVIEW posts before submitting. This helps my host server to save your comments.