Sam Altman says superintelligence may not unlock all science just by thinking harder about existing data Big breakthroughs will still need new instruments, experiments, and machines. "the real world is slow and messy" Progress will come -- but building the tools will take time