Fair for all - DeepMind: The Podcast (S2, Ep8)

AI needs to benefit everyone, not just those who build it. But fulfilling this promise requires careful thought before new technologies are built and released into the world. In this episode, Hannah delves into some of the most pressing and difficult ethical and social questions surrounding AI today. She explores complex issues like racial and gender bias and the misuse of AI technologies, and hears why diversity and representation is vital for building technology that works for all.
For questions or feedback on the series, message us on Twitter @DeepMind or email [e-mail].
Interviewees: DeepMind’s Sasha Brown, William Isaac, Shakir Mohamed, Kevin Mckee & Obum Ekeke
Credits
Presenter: Hannah Fry
Series Producer: Dan Hardoon
Production support: Jill Achineku
Sounds design: Emma Barnaby
Music composition: Eleni Shaw
Sound Engineer: Nigel Appleton
Editor: David Prest
Commissioned by DeepMind
Thank you to everyone who made this season possible!
Further reading:
What a machine learning tool that turns Obama white can (and can’t) tell us about AI bias, The Verge: www.theverge.com/21298762/fac...
Tuskegee Syphilis Study, Wikipedia: en.wikipedia.org/wiki/Tuskege...
Ethics & Society, DeepMind: deepmind.com/about/ethics-and...
Row over AI that 'identifies gay faces', BBC: www.bbc.co.uk/news/technology...
The Trevor Project: www.thetrevorproject.org/
AI takes root, helping farmers identify diseased plants, Google: www.blog.google/technology/ai...
How Can You Use Technology to Support a Culture of Inclusion and Diversity?, myHRfuture: www.myhrfuture.com/blog/2019/...
Scholarships at DeepMind: www.deepmind.com/scholarships
AI, Ain’t I a Woman? Joy Buolamwini, UA-cam: • AI, Ain't I A Woman? -...
How to be Human in the Age of the Machine, Hannah Fry: royalsociety.org/grants-schem...
Find Seasons 1 & 2 on UA-cam: dpmd.ai/3geDPmL
Or search “DeepMind: The Podcast” and subscribe on your favourite podcast app:
Apple Podcasts: dpmd.ai/2Rzlmcu
Google Podcasts: dpmd.ai/3geDjp5
Spotify: dpmd.ai/3w29cb4
Pocket Casts: pca.st/30m1

[@propotkunin445]

In my opinion, the biggest problem with AI isn't AI misinterpreting tasks given by humans (this is a problem, but a solvable one); nor is it AI becoming conscious and rebelling against us; the biggest danger is AI actually doing exactly what certain humans want it to do.
Companies could use it to manipulate potential costumor way more effitiently than via advertising, governments could use it for highly effective propaganga, surveillance, warfare... the possibilities are endless.
And I believe the only way to minimize the risk of too much harm being caused by AI in the wrong hands is social change. If power in governments and companies are structured in different ways than today or the people in charge hold other values because of social change, it'd reduce the risk.
But for that, we probably need way more funding for research in sociology and psychology on social change.

[@texturelessidea]

Maybe I'm just way too naive, but when I found this video I was hoping for a discussion on how Google was going to work towards keeping things involving AI fair between Google and everybody who isn't a share holder of Google. Of course the talk would focus on broader societal issues to distract from the problems directly and intentional caused by Google.

[@thehappycoder3760]

They aren’t mistakes at best they are ignorance, or worse, indifference.

[@kashatnick]

This talk is GPT3 on easy mode.

[@idiotbassist6423]

It's refreshing to hear people who actually know about AI and machine learning talk about these issues.

[@churblesfurbles]

That's all they talk about, as the patterns in the data fail to conform to ideology they obfuscate and cope. The concern for bias is self serving, noted biases of certain sides are never dealt with, for instance democrats vastly overestimate the incidence of police altercations involving unarmed people of certain demographics. The clear uniform bias of those involved make their concerns untrustworthy by default.

[@zweivessaliuz4402]

@@churblesfurbles I prefer favour>bias, but anyways, thank you for your clarification & enlightenment.

[@bobedwards8896]

the world is more than big data...

[@criticalanalyst7546]

great podcasts. i have been learning ML for quite a time and this podcast is filling me with enormous motivation and enthusiasm. thanks Fry

[@thehappycoder3760]

So good, thank you

[@researcher6980]

Godspeed DeepMind ⚡💫🙌💪

[@MsJeannieKellyTX]

☝🏼

[@smokey7927]

Interesting...

[@DM-qm5sc]

And unsubscribed, thanks!