Live Labs online group, computer scientist Gary William Flake was eager to release details of the initiative last week — in one case perhaps more so than others at the company wanted. Flake’s introductory post for the Live Labs Web site publicly rev…

But I do see Google moving rapidly, if somewhat haphazardly, beyond its core business of Internet search in ways that put the Mountain View company on a collision course with other tech giants — specifically Apple and Microsoft. It’s my job as a colu…

But I do see Google moving rapidly, if somewhat haphazardly, beyond its core business of Internet search in ways that put the Mountain View company on a collision course with other tech giants — specifically Apple and Microsoft. It’s my job as a colu…

I see Google’s hubris here as the consequence of a key failing of Information Theory which all computer scientists (including, do doubt, the Google founders) are currently taught. This theory, due originally to Claude Shannon, explicitly ignores the semantics of messages. Information is analyzed in terms of bit streams (sequences of 0s and 1s) and the theory constructed regardless of what these binary digits represent.

This is well and good as far as it goes, and the evidence is all around us of just how far it can go. But, as we enter an age of intelligent computation, and of interaction between intelligent entities (humans and/or machines), then communications is more often about actions than merely the sharing of information. Here, we need to have a handle on the context, the intentions, the meanings, and the consequences of messages, not just their syntactic form. Current information theory, by ignoring semantics and pragmatics, is singularly ill-positioned to provide this handle.