We’re working on details on AT&T Co’s experimental optical computer (CI No 1,353) but in the meantime, the comment from IBM director of research Jim McGroddy to the Wall Street Journal sounds disconcertingly like some of those famous last words that come back to haunt the speaker: It’s easy to get seduced by the idea of computing with light, he says – the work is interesting, the physics is interesting, and their ideas may be useful for certain purposes, but one shouldn’t kid oneself about this area’s applicability, adding that light is good for transmission and connections but not much else; the paper notes that IBM did extensive research on photonics in the 1960s and concluded that computing with light requires more energy than computing with electricity, and that photonics would therefore never compete with electronics.
