In the context of AI, accountability is a question mark

0:12Welcome to the latest FD Advisor Editors podcast with me, Simini Kiriakou. As you may have noticed, we've now branded our Editors podcast as Margin Notes. and when it comes to rebranding we do think of new colours and logos and sometimes new fonts something journalists get very excited about but sometimes branding goes much deeper it could be a complete change in how a company seeks to represent itself and its services the same goes for personal rebrands and sometimes circumstances force us to change our course of direction or even change ourselves. For our guest today, a serious accident led to a transformation into a motivational speaker, disability rights advocate and a campaigner for change.

0:57She's described herself as a bionic woman. She's used her voice to try to make London a safer place to travel, both for people with disabilities and with none, and encourages companies to think more creatively about ways to innovate and design products and services that make things easier for everyone. So joining us is Sarah Delagarde, brighter future strategist for Janice Henderson. Sarah, welcome to Margin Notes. Hi. Thank you for having me. No, it's great. Thank you very much for joining us here. Now, Sarah, you sometimes jovially describe yourself as 80 % human and 20 % robot. Can you briefly explain why you chose that rebrand?

1:34Well, I think that's representative now. I really see myself as being a hybrid being now, 80 % human and 20 % bionic. I wear a prosthetic arm as I lost my right arm above the elbow and I've got a prosthetic leg as I lost my leg below the knee. And can you tell very briefly about what happened to you that saw you lose both your arm and your leg? Yeah, so back in 2022, it was a Friday. I was on my way home from work on the 30th of September and I slipped on a wet and uneven platform and fell through the gap in between a stationary train and the platform through the gap and the train departed, crushed my right arm and then a second train came into the station and crushed my right leg.

2:36There was no one there to hear my cries for help. No one witnessed my full CCTV footage, isn't watched live, there was no staff. The drivers didn't spot me as I lay on the tracks and detected for so long. Even when you were wearing a really bright red or bright pink coat or something? I had the brightest pink coat that you can imagine, and I have light hair, and yet I remained unseen. Luckily, I stayed conscious throughout, and I was able to shout for help for a very long time. And eventually somebody picked up my cries for help and raised the alarm. But it took a very long time for the emergency services to extract me from underneath the train and get me to a hospital.

3:27Yeah, every time I hear that story, it just sort of fills me with absolute horror. It's just so distressing. And I know that since then, you've been really campaigning for better safety, better rail safety. And you've also just very recently won the right to take your case against the London Underground to trial. Can you very briefly talk about that? So I thought at the beginning that I was very unlucky, one in a million. But actually, I discovered that this happens way more often than we think. And so, you know, according to TFL's own numbers, it happens 162 times a month. And for me, that is completely unacceptable.

4:11I mean, I have children who are of age to take the London Underground and I'm terrified every time because I know that safety isn't up to par. And the fact that, you know, there is no staff. We have since also discovered that the help points are not manned. So you remember the story about the lady who got assaulted on the train. She wanted to call for help and none of the help points were manned. And so to me, that is unacceptable. We pay the highest price for that ticket. We pay for it twice, actually. We pay once in, you know, taxes. It's a public service. And then we pay for the ticket when we use the service.

4:52And for that price, we deserve a service that keeps us safe and that gets us from A to B. And I think that's not happening. And people have described the London Underground as being dirty and dangerous. And, you know, I think we deserve a lot better. Absolutely. We do deserve a lot better. And this case that you're taking to trial, I think the date's been set for 2027, five years after the accident. That hopefully will sort of set the path for change and transformation. And that's so important. But in this, you know, I've spoken to you before about this, this whole transformation of yourself has led to such great resilience.

5:38Can you talk a little bit about resilience and how to build up that resilience? That is a really good question. Do you know what? I didn't think of myself as being that resilient. It comes almost as a surprise when you are put in a situation where you don't have much choice but to survive. But I would say that what really helped me the most here was the support I got from others. so I describe it as running a marathon not that I did or ever will but well I don't know you've climbed mountains Sarah maybe you will run a marathon and you reach that point where you think you can't go on anymore so apparently it's called to hit the wall at that stage and it is the people on the sidelines that cheer you on that motivate you to find that strength within yourself that you thought you didn't have And that, to me, is so important.

6:33That's where the resilience comes from. So it's part you thinking, well, I have the choice. I can either let this define me and let depression take over. And it's very tempting, I have to say. It happens on a daily basis. Every morning when I wake up, I have this moment where I think, okay, this is awful. This is not just a nightmare. This is real. I still am missing my two limbs. and there is a moment where you want to lean into the darkness and think oh well what's the point of all of this anymore but then you also have the choice to not lean into that and think well actually I am grateful that I'm alive I'm you know at home safe my daughters my cats my husband they're all there to support me and so that active choice of your mindset that's where that resilience comes from when you are managing it.

7:29But then it is really quite helpful to have people around you who believe in you, who support you, who push you beyond what you think you're capable of. And then you've spoken in the past as well about how Janice Henderson sort of helped you back into work and made things more accessible for you. Could you sort of talk a little bit about that and how important it is for employers and people in your sort of wider sort of professional network to actually make these small changes that can make a big difference? I was so positively impressed by how my employer Janice Henderson had reacted to this.

8:08It started off with, you know, an outpouring of support, both psychologically and financially. You know, people donated to my fundraiser to be able to afford the prosthetics, which was amazing. And really my colleagues showed up. And when I say this, they literally showed up in the hospital by my hospital bed wishing me well. I know that my CFO had traveled from office to office and every time had recorded a small video from my colleagues in Tokyo wishing me well in Sydney and all these different places. and that really made me feel like I could come back. And so, I mean, a lot of people were quite surprised, including myself, that I came back to work after four months.

9:00Four months is not a very long time and I would have never returned to a workplace that was toxic. I only returned because I felt supported and it was good for my mental health to return and to feel useful, to feel like I could contribute in some form, much better than staying at home in my bed and feel really depressed and having worked all my life all of a sudden become a burden. Yes, yeah. Not that you ever want to think about that. You don't want to think of yourself in terms of being a burden, But, you know, people do often find they don't have that support network. And I mean, particularly for you as well, there was, you know, we talk about the big rebrand that's happening, the big reset in a way with everyone using AI and how that's going to touch everything.

9:53But actually, AI has been really helpful to you. You use quite a lot of tools now, don't you, that are AI? Yeah, it was a steep learning curve. I have to admit beforehand I wasn't that tech savvy, but the fact that I discovered that prosthetic limbs have come such a long way. I mean, I was still somehow stuck with childhood memories about pirates and wooden legs and hooks for hands, but actually medical assistive technology has leapt forward significantly and part of it is because of the use of AI. So the way that my bionic arm works is that it is my electric technology. It has sensors, electrodes placed on the surface of my skin.

10:39And so I can think of a movement and move my remaining tendons and muscles inside my residual limb that then gets picked up by these sensors and gestures happen as a result. That's amazing. But that isn't the AI part. The AI part is even better than that because every data point gets recorded. So every time I use the ARM, data gets collected onto an external server. And an AI algorithm sifts through those data sets and starts recognizing patterns. In essence, the ARM learns how I use it. and then once there is sufficient data available it switches over to generative AI algorithms so then it starts predicting what my next move is going to be and writing its own prompts exactly and so what happens then is that the gesture happens more easily for me because before when I first started using the arm I really needed to think it would use a lot of brain power to execute a movement But now the arm knows that when I give it a prompt, it completes the gesture for me.

11:56So very useful in real life because it reduces the time from me thinking about a movement and executing a movement from about 10 seconds to none. 10 seconds seems like a very short time, but when you're trying to pick up a glass, it can be quite frustrating. Answer a call. Yeah, exactly. Hold the cats back from jumping on the counter. Yeah, absolutely. Exactly. But it does bring up this question around AI and accountability. As you know, I'm quite big on accountability. That's why I'm pursuing TFL for their lack of accountability because for TFL to deny all liability means that they don't think anything wrong had happened.

12:47Therefore, they have made no changes, right? And so meaning that other people can go through the nightmare that I'm going through. So accountability is really, really quite important. Now, in the context of AI, accountability is a question mark. And the best way to illustrate this was I presented at my daughter's school to a whole bunch of 7 to 11-year-olds. And the first question I got after finishing my talk was, so miss, if somebody hacks into your arm remotely and your arm then picks up a gun and shoots somebody, who is responsible? That came from a junior or a schooler. My gosh. I know. Good thinking, though.

13:41So, and the answer isn't that straightforward because is it me because I'm wearing the arm? Is it the arm? Is it the companies that produce the different components? Is it the net, the external server? Is it the person who hacked it? Do you have agency in that? Do you? That's the question. And so, I think it's a great illustration to what the danger is with AI. is that it kind of takes away the accountability. But there are obviously things that we can do to keep things safe, keep data safe. Do you think that generally in financial services or generally as human beings, are we that aware of our own personal accountability and keeping everything safe?

14:33I mean, I know we have to keep doing these questionnaires online and the FT keeps sending out, are your data secure? we have to do these little tests and quizzes, but are we really that data aware? I don't know at this point, you know, but what I do think is that AI is extremely powerful and in my mind very similar in that sense to nuclear power. So in the hands of somebody who wants to do good, extremely useful, but in the hands of somebody who doesn't necessarily have good intentions, then it could be world-destroying. Or the careless. Yes, or the callers. Yeah, good point. And so I think that having a regulator for AI developments would be quite useful, and one that isn't just confined to one country, but that is global, because AI doesn't know any borders.

15:25Exactly. I think the European Union has an AI act, but we don't have one yet in the UK. In the US, each individual state has its own sort of AI rules and regulations, but there's nothing cohesive even in the US, which is the birthplace of AI. So how should companies then sort of react to this? I mean, should they just put the barriers up and say, do you know what, actually, the risks are too big for me to get involved? Or should they use it? I mean, if you put your barriers up, you wouldn't be using your prosthetic arm, right? Exactly. And so I think we still need to have an ethical framework ourselves so that we still have critical thinking where you can spot a hallucination and a wrong turn.

16:10And that overall, our aim should always be to safeguard us as humans and not always like I use I use ChatGPT, for example, as a tool. but every time there is an output I want to apply critical thinking to ensure that I'm not completely relying on it because that yeah that that would be to me um dangerous right if you start believing everything without any critical thinking I like the idea of having an ethical use framework in place at every single firm and with every single individual as well to think how am I using AI how am I checking it? How am I verifying it before anything goes out into the public domain?

16:53And in keeping your own data safe, right? I mean, I take it you have safeguards against anyone hacking into your arm. Yes. What did you tell to that 11-year-old? I thought they would have a great career in critical thinking for sure. Yeah, but I did I did respond by saying this arm doesn't function without my input. So it needs my physical input to actually execute something. So the question isn't around somebody hacking remotely wouldn't be able to make it work because it needs my physical input. And so I can also take it off and put it on the floor and it won't run off by itself. Yeah. Well, that is the stuff of science fiction, I guess.

17:42But that is also the fear that people are not sure if they can control AI. And we've seen recently how AI actually said no to a human inputter. And another AI completely wrote another AI, which wrote another AI, it can start writing its own self. So there is a question again of how much do we want to control the prompts? How do we control it? And how do we keep things secure? So do you have the answer to that? How do we keep everything secure? No, I don't think I've got the actual knowledge to be able to put these safeguards in place. But I think sensibilizing people to the issue is just as important.

18:23The fact that you don't want to have biased input into the learning, the training of AI. So if you only get input that is biased towards gender, for example, then the output is going to be biased towards gender too. And so that is something that we need to be vigilant about or knowledgeable enough to be able to say, well, actually, you know, we want the input to be just as diversified as we want the output to be. So maybe we should all be 80 % human and 20 % AI when it comes to our business. I don't wish that upon you, but taking an interest is probably, educating yourself on the topic is probably just as important.

19:12Not just blindly following or just thinking that the computer knows best. That's dangerous, I think. Absolutely. Well, I think we're running out of time. I can see people waiting in the wings. so just for now I want to say thank you so much to all of you, 100 % of you for coming in and sharing a bit of your story and your campaign and talking to us about AI and the importance of safety and safeguards and thank you also for listening. For more news and views visit ftadvisor.com and until next time, please take care

20:24We'll see you next time. automate fulfillment and rate shopping to ship up to 15 times more with up to 90 % savings go to shipstation.com and use code audio to try shipstation free for 60 days

20:39Sarah de Lagarde: some days you just want to amplify your everyday look like when you want the look of false lashes without the extra effort reach for thrive cosmetics liquid lash volumizer mascara or when you want all eyes on your smile keep empower gloss ultra glossy lip serum in your bag it's a burst of 24-hour hydration that smooths like a serum, shines like a gloss, and can be worn sheer or layered. You'll always look and feel your best with Thrive Cosmetics. Plus, every product is 100 % vegan, cruelty-free, and made with clean, skin-loving ingredients that work with your skin, not against it. Amplify your everyday.

21:12Sarah de Lagarde: Go to thrivecosmetics.com slash shine26 for an exclusive offer of 20 % off your first order. That's Thrive Cosmetics, C-A-U-S-E-M-E-T-I-C-S dot com slash shine26.