Journal Articles
Browse in : |
All
> Journals
> CVu
> 292
(9)
|
Note: when you create a new publication type, the articles module will automatically use the templates user-display-[publicationtype].xt and user-summary-[publicationtype].xt. If those templates do not exist when you try to preview or display a new article, you'll get this warning :-) Please place your own templates in themes/yourtheme/modules/articles . The templates will get the extension .xt there.
Title: ACCU Oxford – 28 March 2017
Author: Martin Moene
Date: 02 May 2017 09:24:32 +01:00 or Tue, 02 May 2017 09:24:32 +01:00
Summary: Frances Buontempo talked about Actual Intelligence, and a few people wrote about the evening.
Body:
I really appreciated Fran travelling from London to give this thought provoking talk to this well attended ACCU Oxford monthly meeting. This focused on the philosophical issues behind AI and what intelligence is. One of the standout points for me was about how AI systems are reflecting our prejudices even though this was not the designer’s intention. This is something we will have to guard against, and this will become more pressing as trust builds in AI systems e.g. in the areas of law. I can foresee governments introducing regulations regarding AI system behaviours because what is unacceptable behaviour for humans will be unacceptable for AI systems too. The challenge will be finding unprejudiced data to train the systems on. There was a very interesting observation regarding Microsoft’s controversial AI chatterbot Tay which was if it had been left running would it have cleaned up its act?
Nigel Lester
I was so pleased to catch the great Fran Buontempo at ACCU Oxford, especially so as I will miss Conference.
Fran took us on a quick one hour trip around all of the significant features of the field of Artificial Intelligence, through the lens of a mathematician, scientist and coder, by asking the question, “What is Actual Intelligence?â€
She explained that most of the current rash of successes for ‘AI’ are really just maths.
Using the fantastic Dead Salmon Study https://blogs.scientificamerican.com/scicurious-brain/ignobel-prize-in-neuroscience-the-dead-salmon-study/ she reminded us that statistics and science are hard.
The real challenge for machines and people is kindness. The fact that AI is now raising questions about what it is to be human and how to behave towards others in a way which includes non-humans is an achievement and leads to hope for benefits from the field.
The lively conversation afterwards had memorable contributions from Francis Glassborow, founding elder of ACCU, who pointed out that Human Intelligence has been responsible for much death throughout history.
The evening was rounded off with an explanation of how easy and pleasant it is to contribute to the ACCU publications from the CVu Editor, Steve Love.
Much thanks to Nigel Lister for organising a very well attended evening.
Tim Pizey
I don’t often get a chance to attend the Oxford meetings of ACCU even though I live in the town because they clash with my Bridge teaching commitments. So it was a pleasure to be able to come to listen to Fran’s run through of her keynote for the coming Conference. I probably had far too much to say at the end but her talk was thought provoking (as keynotes should be).
One thing I noted was that less than half the audience had read Gödel, Escher, Bach: An Eternal Golden Braid by Douglas R. Hofstadter. I think this book is essential reading for anyone with an interest in AI. I remember struggling as I approached halfway and wondered if it was time to put it aside and let my mind digest what I had read so far before re-reading the first part and then going on. (I find that is a great technique for dealing with technical books at the boundaries of my understanding.) However I actually managed to keep going and found the second half easier.
Intelligence is emergent behaviour and that whether we like it or not the current trend in computing will result in an alternative intelligence to ours. What I am also pretty sure is that we will not create an intelligent machine through our skills but through accident. I hope it is a benign intelligence because anything like our own intelligence is too dark to want (think of the mess we are making of our world). Of course that leads to the thought that a sane machine intelligence would probably want to remove at least 90% of the human race in the interests of its own self preservation.
Thanks, Fran, for a thought provoking talk.
Francis Glassborow
Notes:
More fields may be available via dynamicdata ..