Is all intelligence created equal? Day 2 of the 'Varieties of Mind' conference at Cambridge
*Note: the blog post will be updated with some useful hyperlinks for further references next week.
I firstly stumbled upon the ‘Varieties of Mind’ conference throughJoanna Bryson’s Twitter profile. Joanna is a leading expert in AI ethics and machine intelligence from the University of Bath. She advertised that there were some tickets still available for this conference held at the University of Cambridge. I booked my ticket immediately after realising that it was organised by the Leverhulme Centre for the Future of Intelligence and having seen the incredible line-up of speakers.
I arrived to the event in my classic fashion - late. The venue was chosen to be Keynes Library at the Cambridge Students' Union. As expected, the interior contains old furniture, portraits of famous alumni and has a gloriously stuffy smell of centuries of intellectual pursuits.
First on the programme was a keynote by Susan Schneider, Associate Professor of Philosophy and Cognitive Science at the University of Connecticut. Susan delivered a memorable presentation and did a great job in unpacking some complex concepts such as 'consciousness'.
She highlighted that it is neurologically possible to create consciousness in a machine. However, as we develop conscious AI, we have to device better tests for it too. Interestingly, a sophisticated but not conscious AI could trick a human tester into thinking that it is indeed conscious.
Next up was an amazing panel/series of talks by Sabine Hauert from the University of Bristol, Jose Hernandez-Orallo from UP Valencia, and Yasuo Kuniyoshi from the University of Tokyo.
Sabine’s presentation was focused on the concept of swarm intelligence and its applications. In the animal world, it is represented by bees, birds and ants. In terms of our study and potential recreation of swarm intelligence, two distinct approaches are ‘bio-inspiration’ and ‘exploration’ (see the slide below).
I was happy to learn that swarm intelligence-based particles can tackle cancerous tumours! Another interesting discovery is the fact that some robots in Bristol Robotics can grow limbs over time (albeit very basic ones). I asked a question about action films and my observation that whenever there is an evil army - the group of 'good guys' decide that to defeat it they have to take out the ‘queen’ of the hive/swarm. Sabine clarified that there wouldn’t necessarily be a ‘queen’ in swarm intelligence. Although the swarms vary in the degree of centralisation of control systems.
varying types of intelligence that currently exist. Artificial (machine) intelligencenon-human animals have different set of skills depending on the species; humansgenerality and capability Next speaker, Jose, described the is extremely good in a narrow set of tasks; are reasonably good at various tasks requiring cognitive input. I also enjoyed the graphically represented (second photo below) relationship between the of different types intelligence - great visual comparison!
The last speaker of this panel, Yasuo, showed some of the fascinating advances in AI that he had been involved in. He was specifically addressing the need to investigate very early human development to improve our understanding of cognition. I was especially amazed (and slightly disturbed) by a computer simulation model of a human foetus. It strongly resembles human behaviour and adapts to the environment.
After the first panel, delegates were invited to have a tea and cake break. Needless to say, this was met with an overwhelming excitement among the attendees.
Somewhat surprisingly to me, I found the topics raised by the second panel incredibly interesting! Jonathan Birch (LSE), Colin Klein (ANU) and Henry Shevlin (Cambridge) were discussing sentience and suffering among species. Check out the pictures of some of the slides, amazing stuff! I was particularly impressed by how much research and thinking is underway on topics such as dimensions of pain among species, animal sentience, and how to reconcile suffering and pain.
The day was concluded by both educational and entertaining debate that took place in the grand Debate Chamber at Cambridge Union. The question of the debate was whether humans can understand non-human minds. Kanta Dihal and Jack Lyons argued that we CANNOT, while Sarah Darwin and Charles Pigott argued that we CAN. The CANNOT’s won by a small margin but it was a super stimulating debate - some videos of the debate will be linked up shortly!
Huge thank you for reading!