#huxleysummit - @joswinson argues for a code of ethics for #AI. I absolutely agree. But how do we hold people to it? Last week's #CUKC roundtable tackled this. Conclusion? No conclusion. Because competing interests of business, academia, and law-makers mean it's hugely difficult.
There are several approaches: I'd love to see an external ethics review conducted for any project (academic or commercial) using human participants or human data. Academia already does this but businesses are reluctant as their research is confidential/sensitive. #huxleysummit
We could try an ethics charter along the lines of equality charter/Athena SWAN, where businesses/research groups voluntarily sign up to a code of conduct. But, as @julianhuppert said at #CUKC, who funds this? Who evaluates it? #HuxleySummit
Look at the Facebook fiasco as an example of the difficulties of legislation across jurisdictional boundaries, and when dealing with corporations where any monetary fine you impose is only a tiny fraction of their daily revenue. #HuxleySummit
A 'hippocratic oath' for #AI is well and good, but who is writing it and who is policing it? #HuxleySummit
(As a little aside: we don't even have a universal human code of ethics, let alone one for #AI. No biggie, eh?)
There are LOADS of suggested AI ethical frameworks out there already. As @BVLSingler says, do we really need another one? #HuxleySummit
Just to reassure you all, @BVLSingler and I are up here tweeting furiously and sorting this all out via 280char rants. The future is in our hands and we should be done by about tea time, so get the wine open.
