|
![]() |
Homepage / Publications & Opinion / Archive / Interviews (pre 2002)![]() Sci-fi or sci-future? Hollywood technology scrutinised Thursday 18th October 2001, 7:15am, Joey Gardiner Cochrane and Kurzweil on future high-tech, as seen in the movies... We asked two of the world's top tech thinkers to cast judgement on concepts seen in some well-known films. From memory implants in Johnny Mnemonic to the paranoid ship's computer in 2001, from uploading a virus to aliens in Independence Day to universal translators in Star Trek, we asked for advice from Peter Cochrane, entrepreneur and former CTO at BT, and Ray Kurzweil, inventor and futurologist. Here they explore the credibility of some of the enduring scenes Hollywood has churned out. How credible are the ideas on show, when might they become reality (if at all), and what do we make do with for now? Read on, and feel free to add a Reader Comment of your own... JOHNNY MNEMONIC PETER COCHRANE: People have been using their brains as bit transport for millennia - we hear, see, feel, smell, and convey to others via language and pictures. I think the real contest is with electronics which look set to exceed our mental and storage capacities within the next 50 years. Chip implants for hearing are now common, chips for motor and sight restoration are at an experimental stage and it is only a matter of time before we move onto memory augmentation. But today our technology looks like smoke signals trying to communicate with a mobile phone. RAY KURZWEIL: Downloading and/or uploading information to/from the brain will require substantial integration of biological and non-biological intelligence. We've started down this road with contemporary neural implants. Ubiquitous and non-invasive brain implants using 'nanobots' (robots the size of blood cells that will travel through the capillaries of our brains and communicate with our biological neurons) will start around 2030, and will be deeply integrated with our thinking process in the 2040s. 2001 PETER COCHRANE: The principal reason our machines are failing to develop true intelligence is that they are deprived of all sensory input. Today we cannot define, describe, quantify or measure life, intelligence or emotion. Until we can, discussion is mostly a waste of time. We may not be smart enough to get there without machine help - but they may become smart enough to do it without ours! RAY KURZWEIL: Although HAL was paranoid (a not uncommon human condition), he operated at fully human levels of intelligence. He could pass a Turing Test which requires being able to respond to dialogues in a way that is indistinguishable from humans. I expect that machines will pass the Turing Test by 2029. INDEPENDENCE DAY PETER COCHRANE: This is the here and now - it works! And it comes in numerous sophisticated forms e.g. 1) Wipe the memory clean now or at some pre-defined trigger point, 2) Corrupt memory so that errors are imperceptible to the user but significant for you, 3) Change the software to do your bidding now or at some later date. RAY KURZWEIL: Information warfare will grow in its importance. I put 40 years because we are not yet in a position to completely disable an enemy through software pathogens (e.g. viruses). This would require our enemies to be largely dependent on software-based intelligence. However, prior to these scenarios, we will be greatly concerned with biological terrorism and warfare, and in particular with biological pathogens that are genetically altered. This involves the combination of information technology and biological knowledge. STAR TREK PETER COCHRANE: This works right now at a very basic level for text and speech - but not at the level of natural language spoken interchange. Computer speech recognisers are better than humans today on a word-by-word basis, but they lack cognition and contextualisation. Until machines become fully sentient we will not be able to realise the Babel Fish. Look at the trouble human translators have. Ultimately I think machines will be better but will fall short of the ideal Babel Fish. RAY KURZWEIL: We have speech recognition, language translation, and synthetic speech today. The quality is far from perfect, but getting better every year. The quality demonstrated in these movies requires human level intelligence, which I expect by the year 2029. BLADERUNNER PETER COCHRANE: This is only likely to be a matter of time. My guess is that once a machine is truly sentient, connected to our world through adequate sensors, with fully adaptive hardware and software, with an evolutionary (replication/reproductive) capacity, it will follow naturally. All animals show emotion of varying degrees and we can expect machines to follow. There appears to be nothing special about carbon life - it is just that we have never seen any other forms. RAY KURZWEIL: Our ability to recognise and respond appropriately to emotion is part of our intelligence. Indeed, it is the most complex and most intelligent thing that we do. Mastering emotion by machines will require machines with human-level intelligence, which will occur by 2029. Peter Cochrane is the co-founder of ConceptLabs and spends his time around the world getting innovative start-ups off the ground and evangelising technology, among other things. He is well known as the former head of BT Labs, Martlesham, when he was CTO at the telco, and is a writer on various technology issues. Ray Kurzweil is the founder, chairman and CEO of Kurzweil Technologies, known for a print-to-speech reading machine for the blind as much for his piano synthesiser and other breakthroughs. He wrote The Age of Intelligent Machines while at MIT and was awarded the 1999 National Medal of Technology by Bill Clinton. |
![]() |
||
![]() |
|