Technology will always rush away because policy too often doesn’t anticipate; it responds. It goes through cycles of refinement, and it needs to be right.Danil MikhailovExecutive Director of data.org
Perry: That’s true if you’re a politician, but it’s also true if you’re a CEO sitting next to a startup founder, or a media professional writing a headline. New technology is always the sexiest thing on the block. You want to feel connected, and you want to have something to tell your board when they ask, “What are you doing about ChatGPT?” Being cognizant of new technologies and how they will affect your organization is critically important, but I think being measured, rather than falling madly in love, will serve you very well.
Danil: The solution is to look at other sectors where we’ve faced similar problems. In pharma, drug development can be fast, but we want a legal framework in place for how you test a drug’s effectiveness through multiple cycles before you release it to the general population. I think we can look at the health or life sciences sectors to see what could be learned, and what can be applied, in terms of regulating AI.
Perry: And I think, while new technologies are sexy, they’re most effective when more diverse people are trained in using them. So invest in people – in women, in more diverse geographies and communities – to improve design, implementation, and outcomes, and reduce unintended consequences. That will lead to really important high-impact proliferation.
Danil: I hope that we’ll also use this moment to reimagine how we train young people, particularly in data science and technology skills. That means broadening the definitions of those fields away from just the technical perspective, and towards creating people who are more interdisciplinary and more multilingual in terms of their skills. They should understand how to work with vulnerable communities, and understand the social consequences of design. Our STEM courses should incorporate the humanities and the social sciences, from middle school to high school to university and right through professional training. Those already in their careers should take this transitional moment as a chance to retrain as well.
Perry: That’s why one of our goals is to build 1 million purpose-driven data practitioners by 2032. We understand that, fundamentally, change is about people. We need people who understand the nuances of the data.
Data collected by humans has ethical and personal implications, and that reflects the need for a new kind of data – data for social impact.Perry HewittChief Marketing and Product Officer at data.org
Danil: For me, AI and data science is a social technology. AI, fundamentally, is trained on data that comes from humans, and it is fed into systems that affect humans. Therefore, on either side, humans are integral, and society is integral. That means that the experts making decisions must have as much training in the social sciences and humanities as they do in tech. One thing we’re doing at data.org is defining the missing career roles and skillsets required to safely and responsibly deploy AI in society. In particular, we’re seeking out the roles that will make balanced decisions when considering and designing for the whole ecosystem. To use the example of the design of cities: we have builders to create individual houses, but we also have city planners to design the whole city in a way that balances private spaces with public, and commerce with recreation. In data science, engineers would be the equivalent of builders – but where are the city planners in AI and data? That’s Perry’s question, by the way.
Perry: Yes, and one of the challenges is that local and global contexts differ so dramatically. People are, understandably, skeptical of the importance of AI in regions that don’t have internet access, or even electricity. The uneven distribution is a risk, and I hope that the private use of AI will be so universal that there’ll be government representatives and private and social representatives invested in leveling the playing field so that many more have access to these tools. But we in the philanthropic and social sectors need to be very intentional to make that happen.
Danil: Somebody needs to be thinking holistically about the whole ecosystem. Not only do we need to create that role within big companies, but we need to create a whole educational and career pathway for it.
AI’s great power is in acceleration to scale, but there are risks and opportunities in the changes that are coming in the relationship between us and information. The risk is the rise of misinformation – a huge explosion of misinformation that will swamp us all. But if we have safeguards in place to prevent that, and which truly democratize access to these tools for even those who are currently excluded, then the possibilities are immense. In education alone, we could translate all the content of the world into every other language of the world. Wouldn’t that be wonderful?
Perry: That potential that you and I see is, I’d say, our greatest hope. There are tremendous opportunities for the use of AI – in coding, in testing, in marketing, in product development, and in so many more areas which would benefit from AI automation. But one of the biggest challenges with any new technology is euphoria, and I confess that I’ve experienced euphoria myself at the possibilities of this new technology. The biggest risk is that we’re not mindful – that our training data is incomplete, that the people implementing and using it are inadequately diverse, and that we don’t question the results of the machine.
Launched by The Rockefeller Foundation and the Mastercard Center for Inclusive Growth, data.org is a platform that brings together philanthropy, private sector technology, academia, and social impact organizations to build the field of data for social impact.
Danil Mikhailov is the Executive Director of data.org. Prior to data.org, Danil was at Wellcome Trust, where he founded and directed the Wellcome Data Labs. Danil holds a Ph.D. in Sociology and Communications from Brunel University London.
Perry Hewitt is the Chief Marketing and Product Officer at data.org. Perry is also a Connect Council member at Glasswing AI Venture Capital, a Stimson Center Loomis Innovation Council member, and a board member at Solar Sister.
For more information about the work of Perry and Danil, you can visit data.org, and read the Workforce Wanted report which was the foundation of their 2022 Bellagio convening. You can also connect with Perry and Danil on LinkedIn.
Welcome to a special edition of the Bellagio Bulletin, where you’ll have a chance to hear from leading voices within the alumni network on one of the greatest global challenges of our time – the ethical application and governance of artificial intelligence. We hope you’ll find their points of view as illuminating as we have […]More