Another sequence from Doug Hofstadter responding with scepticism to Ray Kuzweil at the Stanford Singularity Summit (2007 or 2009 ?) – singularity as in machine intelligence or machine aided intelligence overtaking human intelligence. Sceptical in terms of short predicted timescales, immortality, time travel, hockey-sticks, etc … losing credibility … but concerned that mainstream science is not taking the core ideas seriously and finding any serious arguments against the plausible possibilities.
Kurzweil, Wolfram, Hameroff, Chalmers all the usual suspects at the 2009 Summit – need to listen to Chalmers and Scmidhuber – this community needs some philosophical injection into all the exponential processing power hype – intelligence and intellect are far more than processing power. (Thanks to Krim over on MD for the report on the 2010 H+ event.)
Schmidhuber is excellent – a German with a sense of humour – as well as the explicit jokes, his take on the psychology of seeing exponential approaches to points of significance is a very clever dig at Kurzweil (my ’twas ever thus meme) – truly excellent talk – compression is key.
Chalmers is also excellent – does philosophy proud – basic logical argument, philosophy of mind and metaphysics. His controlled simulation and gradual upload consciousness and identity cases are pretty much the Hofstadter and Dennett’s “Brain in a Vat” thought experiment.
Interesting even Hameroff, though he mentions microtubules at length – in dendritic gap connectors and in internal neuronal networks – he doesn’t major on quantum coherence – brain as a quantum computer – as a mechanism for distributed coherence of consciousness, until the light-hearted Spielberg-AI moment towards the end. Much of this stuff does indeed seem like valid (testable) science. Though the first question – cut-off immediately by the mediator – brings him back to this unpopular Hameroff-Penrose topic.
Ben Goertzel too, came over well (these are all 2009) … integration … cognitive synergy …. interoperation … lifting & lowering of knowledge too – from communication language to semantic nodes and links of knowledge and back to communication language.
The word that jumps out at me from all of this is Integration. (ie not processing power.)