Ethics isn't a box to check in science and machine learning, it's what makes the work better. Asking ethical questions forces you to understand your data and patients more deeply, which produces more robust solutions, better medicines, and actual trust. The speaker's organization embeds ethics experts directly with engineers instead of treating it as an afterthought.
Chapters
We kind of fundamentally believe that ethical science, right, is better science. And the reason why it's better science is those types of questions force you to engage more. Well, this is the data I've got. Forces you to engage more, forces you to understand how to collect the data, why they have structures to get the data. It gets you involved in the patients' lives. It helps you understand the systems that have that data in a real practical sense, which means you probably build a better solution. right? A better solution, right? We fundamentally will win out in the market. It's probably more robust as well from machine learning perspective that, so, and it can be used more. So it is a, it is fundamentally comes down to trust, right? Making an attempt to build the best thing you can for everybody. And in the process of doing that, you engage more people, you make that, you also might get a flywheel effect, right? You're solving problems, maybe understanding why people don't come into trial. Maybe you can fix something to get the data that now there are more people in trials and we actually can come up with better medicines, right? More representation. So there is a flywheel effect of doing that properly. For us, it's not, as Russ was talking about sprinkling ethics about it, it's not about having the ethics police, right? So too often you see people trying to bring ethicists in on top of the team as a regulatory thing or something else coming after the fact or things like that. You can't do it like, it has to be done at the same time, right? So you have to think very carefully about the ethical culture you create with the machine learning people. And that's something where we are very keen to have, right? Is the right kind of people. So this is why we have responsible AI people and a VP of ethics and policy around these things, 'cause policy and ethics are actually kind of intertwined, right? That's why we have regulators, right? What are the regulators doing, right? They're actually enforcing a lot of the principles that Russ actually stated earlier, right? That's why we have regulators in society and things like that. So there is a constant dialogue between that. And I think, so this is something we've thought very carefully about. And so we have— there is— they're fully fledged part of the team, but they're also technical people as well. And that's a very key thing.
