Mastercard: The Human Equation: Using AI and Data To Address Global Challenges and Build Inclusive Economies - Seite 2
02
Innovating with guardrails
The internet was not based on a single piece of technology. Instead, its enduring power emerged when Bob Kahn and Vint Cerf - known as the "fathers of the Internet" - invented the protocols and architecture that allowed computers to form networks with one another. "As long as the internet adhered to the basic architecture, it could continue to evolve," Kahn said. That approach allowed the internet to persist and expand far beyond its first small network of computers.
For AI to make an impact, it will need a similar set of protocols and architecture to create interoperability on an international level.
Regulation will have a role to play, too. AI should be serving the communities in which it exists, and these needs vary widely depending on location, so regulation cannot be one-size-fits-all; it requires context to be effective. "Technology is extremely difficult to regulate for a number of reasons. It evolves very quickly. What you don't want is a static piece of regulation that is built based solely on the way the technology works today," said Dorothy Chou, head of policy and public engagement at Google DeepMind. "Historically, what we've seen is that good regulation actually creates public trust."
03
Together, we go far
When the COVID-19 pandemic hit, Kenya's government wanted to make informed decisions for the health and safety of its citizens, explained Shikoh Gitau, founder and CEO of Qhala, a Nairobi-based consultancy that specializes in health informatics and the technology of social impact. But every health center had its own small and private data silo. So policymakers were forced to follow frameworks created on other continents, which ultimately proved poorly suited to the needs of Kenya.
Information that's fragmented by unnecessary barriers or delayed by bureaucracy loses its ability to make an impact. The way to change these limitations is through collaboration between the public and private sectors. As Holly Krambeck, director of development data partnership for the World Bank, said, "As much as I hate to admit it, international corporations can't solve everything, so we need global partners of all different types."
Lesen Sie auch
04
Drawing in and opening up
As of now, 85% of all AI developers are men, according to Gabriela Ramos, assistant director-general for the social and human sciences at UNESCO. And, as several panelists noted, the majority of data fed into AI comes from North America. That means AI models are being trained with data about a world from a tiny sliver of the global population. In fact, there is almost no data representing the Global South at all - and if there is, it's outdated and in incompatible formats.