On the front cover of their initial business plan for DeepMind, the AI lab they set up in 2010, Sir Demis Hassabis, Mustafa Suleyman and Shane Legg wrote a single sentence: Build the world’s first artificial general intelligence.
Their view, which remains true today, is that traditional AI technologies were too “narrow”, able to perform brilliantly, but only after humans had laboriously trained them using vast data sets. That made AI excellent at tasks such as analysing spreadsheets or even playing chess. But, artificial “general” intelligence, known as AGI, had the potential to go even further.
Fifteen years on, technology CEOs are united in the belief that AGI is the next big breakthrough and are waxing lyrical about its potential.
您已閱讀4%（741字），剩餘96%（16030字）包含更多重要資訊，訂閱以繼續探索完整內容，並享受更多專屬服務。