Geoffrey Hinton once helped spark the neural network revolution that powers today’s artificial intelligence systems. Now the Nobel laureate sounds repeated alarms about what those systems may become. At a recent Sana AI Summit in New York, the 77-year-old computer scientist told his audience that humanity stands on the edge of something profound. We are not merely designing tools. We are bringing into existence new forms of intelligence that will exceed our own. “I think it’s going to get much more intelligent than us — that’s my guess,” Hinton said . Nobody will beat these systems at Go or chess again. Look at what they already achieve in mathematics. An AI proved one of Paul Erdős’ theorems that very morning, drawing on an entirely new branch of the discipline. The pattern feels clear. In closed domains, these models generate conjectures, test them, learn from setbacks and compound their knowledge without end. The same trajectory now appears in language models.…