‘Solve all diseases,’ you say?

At Google I/O, DeepMind CEO Demis Hassabis said the company aims to reimagine drug discovery with the goal of one day solving all disease. He delivered this with a straight face, like it was just another product feature.

This is classic tech conference rhetoric. Big, vague promises about transforming healthcare that skip over the hard parts like clinical trials, regulatory approval, and the fact that biology is wildly more complex than code.

Google has real AI tools in this space, like AlphaFold for protein structure prediction. That's genuinely useful for researchers. But there's a massive gap between predicting protein shapes and actually curing diseases.

For anyone building with AI, this is a reminder to stay grounded. The technology is powerful for specific problems, but sweeping claims about solving entire fields should raise eyebrows, not excitement.

The pattern here matters more than the specific claim. When companies talk about AI solving everything, they're often selling a vision rather than a product. Keep that in mind when evaluating tools for your own work.