• 9 months ago
In 2011, Kay Firth-Butterfield picked up a copy of TIME titled 2045: The Year Man Becomes Immortal. The trend of exponential growth in computing power, that magazine issue projected, meant that computers would surpass the brainpower of a human in 2023, and of all humanity combined by 2045. Firth-Butterfield realized at that moment that getting artificial intelligence right would be absolutely critical.
Transcript
00:00 [Music]
00:06 Thank you. I want to start by saying thank you to all of you and to Time magazine.
00:14 But I also want to start by thanking my family because without them, and I'm sure it's true for all of you here,
00:23 without our families who support us, and particularly me in my determination that we set out now
00:31 what we humans want to do with our lives with AI, then we wouldn't have the time or space to have these conversations.
00:42 What I said on the film was that I don't want us to wake up in 2050
00:50 and find that we wished we had had conversations today about artificial intelligence
00:58 and how we human beings interact with it for our future. I genuinely believe that we have a great future with AI
01:08 and it's in our hands to shape it now. We need to do that now.
01:16 My story in Responsible AI would not have started without good journalism.
01:23 And good journalism is something we need to protect now because we see deep fakes,
01:32 problems with elections, and journalism is under threat from AI.
01:38 But also it has great potential where AI is used wisely in journalism.
01:46 Back in 2011, I picked up a Time magazine edition which featured a cover story entitled
01:57 "2045, the year man" - I'll come back to that word "man" - "becomes immortal."
02:07 At the time I was writing a book about human rights and I was immediately fascinated by the idea of
02:14 how we as humans would live and work with powerful computers.
02:21 In fact, I was so enthralled that I devoted the final chapter of my book to pondering
02:29 what I thought would be those complex issues of interactions between humanity and machines.
02:38 I was on my path that led me here and hopefully beyond.
02:44 Put simply, we need to work out now if humans want to fly the plane as they do in Star Wars or Star Trek
02:56 or whether we want to hand it over to AI.
03:00 It's a simple analogy for our current and future interactions and a decision we need to make now.
03:10 And that brings me, with planes, to another story.
03:15 Our daughter is a pilot in the US Air Force.
03:20 There are about 6.5% of pilots in the Air Force who are women and only 3% of them fly fighters.
03:30 You remember that Time article I referenced, the year "man becomes immortal."
03:38 This brings me to a point about data, the oxygen of AI.
03:45 Our daughter's data as a female military pilot is overwhelmed by that of her male colleagues,
03:54 whether it comes to flying or, perhaps more importantly, to her health care in the event of an injury.
04:03 If we are to succeed in creating better economic prospects for all with AI,
04:12 we have to start by doing better with data.
04:15 About 3 billion people cannot access the internet.
04:22 Put that against the 180 million or so who use chat GPT monthly.
04:29 Billions more have not created a sufficiently large data footprint for their contributions
04:37 to be evaluated by generative AI without very precise prompts.
04:44 Indeed, the bulk of the data on heart attacks comes from white American men over 55.
04:56 I guess it's probably no surprise to you that the majority of data comes from white men,
05:04 mainly based in the global north.
05:08 It's just that they've had the pen for longest in the data that we are using to train our generative AI.
05:17 It seems sometimes that we haven't come very far.
05:22 Jane Austen wrote her novel Persuasion in 1817.
05:29 Her heroine said, "I'm not going to refer to books to back up my argument about women's feelings,
05:39 because they've all been written by men."
05:43 I hear you saying, "Well, that's just the way the world is."
05:49 And I agree, that's true.
05:51 But one of the power and potential, and also the peril of AI, is that we can now go beyond
06:00 how it is to collect and use data of how we mean it to be.
06:10 We have far to go, but I think in the last year we've seen some great things in terms
06:15 of organizations putting in place good foundations for responsible design, development,
06:24 and use of artificial intelligence before they go ahead with AI solutions.
06:33 I have hope that we can truly come together to make AI a safe and equitable tool that
06:41 we use to advance humanity as well as our economy for everyone.
06:49 So I ask you that we plan together our future and we start doing so today.
07:01 Thank you.

Recommended