I revisited my earlier description of AI as brute-force learning. Later tools changed my own work and forced me to reconsider a judgement formed from earlier systems.
I used AI, decentralised ledgers and nuclear fusion as examples of technical work whose eventual application may differ from its popular story. I defended continued experimentation where results had begun to challenge an earlier dismissal.
I still see a live tension between resisting hype and updating a view when the evidence changes. The technology references remain anchored in 2023.