Should We Fear the Rise of the Machines?

Dr Allan Tucker

When?
Tuesday, January 21 2020 at 7:30PM

Download iCalendar file
(e.g. import to Outlook or Google Calendar)

Where?

38 West Street
Chichester
PO19 1RP

Tel: 01243 783185
Email: info@chichesterinn.co.uk
Website: www.chichesterinn.co.uk

Who?
Dr Allan Tucker

What's the talk about?

Join Chichester Skeptics in the Pub at 19.30 on Tuesday 21st January to hear from Dr Allan Tucker on ‘Artificial Intelligence: Should We Fear the Rise of the Machines?’.


Artificial Intelligence has been a staple of popular culture for generations, from promises of a Leisure Age when robots would do all of our work while we lived in luxury to scare stories where robots turn against their owners. Dr Tucker will look at some of the key figures in AI during the last century through to the state-of-the art learning machines that are all around us today. He will also ask whether we will ever really have intelligent machines that are comparable to humans.

Dr Allan Tucker is a reader in computer science at Brunel University London. His research focuses on designing algorithms that can learn models from data to aid the understanding and prediction of important events (such as disease onset from clinical data and ecological changes in environmental data). He collaborates widely and has current projects with NHS Digital, the Zoological Society of London, and Moorfields Eye Hospital.

All are welcome — we look forward to an engaging and informative evening. Entry is £5.

Summary of talk

Dr Allan Tucker's talk was a great start to our 2020 programme.

Speaking to a well-attended audience, he began by considering the history of Artificial Intelligence (AI), both cultural and scientific, from the automata of the 18th Century to the Alan 
Turing test.

Early post-war advances in AI led to over-optimism. Clever machines and robots would transform society. This is now happening but much later than predicted.

Dr Tucker highlighted some of the most difficult problems: language, motor skills, facial recognition. Great strides were made with increasing computer power but the major breakthrough was creating software that emulated the way the human brain is organised (neural networks).

Even so, Dr Tucker doesn't think we'll have driverless cars in his lifetime. The amazing human brain is the result of billions of years of evolution. It is unrealistic to expect relatively simple computers to be able to perform the complex human interactions involved with driving let alone more sophisticated behaviour.

Dr Tucker used many interesting and often amusing examples to demonstrate the current state of AI, including IBM's Watson taking part in a TV quiz show, Deep Mind playing chess, clinical diagnostic aids, legal implications, AARON creating art, etc.

He spent some time considering the mystery of consciousness and wondered if this was possible for AI, and described some of his early university studies on experiments with brain lesions.

Dr Tucker ended his talk by considering the implications of allowing AI to collect and analyse our personal data when not even the creators of these so-called Black Box systems can ever know how they produce their output.

After the break, Dr Tucker answered many interesting questions, including:

Could machines take over the world? It's possible but unlikely. The greater threat is the way powerful people and organisations make use of AI.

Do you think AI will continue at the rate as it has so far? No but it won't stop altogether. It's still a valuable industry.

Will we need to restructure our economy to stop the risk of AI developing beyond our control? AI will definitely affect the economy but I don't think it will get out of our control.

Should the developers of AI make a greater effort to explain AI to the public? Yes, most definitely.