BBC Radio 4 - The Reith Lectures - Available now https://www.bbc.co.uk/programmes/b00729d9/episodes/player
In December, Start Russell, founder of The Centre for Human Compatible Artificial Intelligence at The University of California, Berkeley gave the four British Broadcasting Company (BBC) Reith lectures on January 3rd, 10th. BBC Radio 4 - The Reith Lectures
"Machines don’t have an IQ. So, to talk about the IQ of a machine doesn’t make sense.”
“In 1936, in his early twenties Alan Turing said: “Once the machine thinking method had started, it would not take long to outstrip our feeble powers. At some stage therefore we should have to expect the machines to take control. Humans are intelligent to the extent that our actions can be expected to achieve our objectives.”
Stuart Russell went on: “We have to build AI systems that know they don’t know the true objective, even though it’s what they must pursue. This would solve the control problem.
“Over the next few days, partly in deference to the Three Laws of Robotics proposed by the great science fiction writer Isaac Asimov, I wrote these ideas down in the form of three principles. The first two essentially say what I just said.
“The first principle is that: The machine is only objective is to maximise the realisation of human preferences.
“The second principle is that: The machine is iinitially uncertain about what those preferences are.
“The third principle is that: The ultimate sourrce of information about human preferences is human behaviour.
“Stepping back a little, I think we have to move from the current situation, where AI researchers think they’re doing good AI, but the ethicists are wagging their fingers and saying “Bad, bad!” to a situation where the AI researcher gets up in the morning and doesn’t just say “Okay, okay, I’m going to listen to those insufferable ethicists today,” but instead, says, “Today I’m going to build a really high-quality AI system.” And what that means is an AI system that’s provably beneficial to humans, just as when a doctor strives to be a good doctor, what that means is healing people and not lining one’s pockets by selling fake medicine."
To me, it seems like every AI centre that’s working ought to have a resident philosopher or ethicist on board to prick the consciences or at least remind people that this is important. Are those people there?”
“STUART RUSSELL: Yes. What works is for a real conversation to happen between AI designers and Ethicists.”
HIFA profile: Richard Fitton is a retired family doctor - GP, British Medical Association. Professional interests: Health literacy, patient partnership of trust and implementation of healthcare with professionals, family and public involvement in the prevention of modern lifestyle diseases, patients using access to professional records to overcome confidentiality barriers to care, patients as part of the policing of the use of their patient data
Email address: richardpeterfitton7 AT gmail.com
