Don’t think about data as the new oil, Oxford economist Carl Benedikt Frey argues in his DLD Summer session with Azeem Azhar – because data alone doesn’t make you a leader in artificial intelligence. What’s really needed, he says, is making algorithms smarter, so that they understand in which context they’re operating.
“There’s a risk of exaggerating China’s AI capabilities”, Frey says, and he sees “a real opportunity for Europe“ because “the AI race is in its early innings.”
Azhar – publisher of the popular Exponential View newsletter – agrees that the current AI approach is limited. But “the reality is that very large models are working”, he says, which benefits technology giants and creates “an entire supply chain” of software and hardware around the dominant machine learning approach of neural networks trained on huge amounts of data.