July 25, 2022 at 9:30 am
This is a fascinating area for me. I enjoy the science fiction that addresses these issues such as Do Androids Dream of Electric Sheep (I think the book is more interesting in this area than the film adaptation, Blade Runner). Westworld, the TV series, and the Hyperion Cantos by Dan Simmons are more contemporary examinations. OK, they're a long way from addressing real issues now but you've got to think ahead 😀
July 25, 2022 at 1:26 pm
A book that addresses real issues is "Weapons of Math Destruction", by Cathy O'Neil. She talks about how AI algorithms build in unfair behavior, and usually in a way that is visible to almost nobody. For example, resume parsers may give extra weight to degrees from certain universities, although doing so skews the demographics and disadvantages most minorities. That's a very simple example, and there are much more insidious ones. Highly recommend this book.
July 25, 2022 at 3:39 pm
Yes, implementing ethical rules into an AI system is a tricky subject, especially when society in general can't objectively define what is ethical or even abide by existing codified ethics laws. But there are some industries like healthcare and insurance where regulations and internal procedural rules govern how they have done business for decades (how should patients be triaged when the emergency rooms are inundated?, what factors should influence insurance rates?, etc.).
Another thing to keep in mind is that ethical decision making involves qualitative analysis. Can an AI really choose which patient gets the kidney transplant when there are so many qualitative and subjective issues to consider?
"Do not seek to follow in the footsteps of the wise. Instead, seek what they sought." - Matsuo Basho
July 25, 2022 at 9:15 pm
I think it's less the AI making a decision in some fields, and more what is their recommendation? We have data for decades, but not flawed/biased/etc is that data? We see a lot of that in fields that use previous data on decisions for training.
July 29, 2022 at 8:48 pm
A book that addresses real issues is "Weapons of Math Destruction", by Cathy O'Neil. She talks about how AI algorithms build in unfair behavior, and usually in a way that is visible to almost nobody. For example, resume parsers may give extra weight to degrees from certain universities, although doing so skews the demographics and disadvantages most minorities. That's a very simple example, and there are much more insidious ones. Highly recommend this book.
That's a perfect example of what I was going to bring up. WHOSE "ethics" will be imparted? And then there's religious and political "ethics". Lordy...
--Jeff Moden
Change is inevitable... Change for the better is not.
August 9, 2022 at 9:43 am
A book that addresses real issues is "Weapons of Math Destruction", by Cathy O'Neil. She talks about how AI algorithms build in unfair behavior, and usually in a way that is visible to almost nobody. For example, resume parsers may give extra weight to degrees from certain universities, although doing so skews the demographics and disadvantages most minorities. That's a very simple example, and there are much more insidious ones. Highly recommend this book.
The real issue here is why the title is not 'Weapons of Maths Destruction' surely? This would be correct English and sound better, i.e. more like mass, too 😉
Viewing 7 posts - 1 through 6 (of 6 total)
You must be logged in to reply to this topic. Login to reply