Interestingly enough, it seems it is the same as we stand today: see gMail, spamassasin, bayes condition. Turing (award), GOOG.
A very senior Microsoft developer who moved to Google told me that Google works and thinks at a higher level of abstraction than Microsoft. "Google uses Bayesian filtering the way Microsoft uses the if statement," he said. That's true. -- http://www.joelonsoftware.com/items/2005/10/17.html
Empathy, Sapience, Self Awareness,Consciousness, Sentience, Abstraction, Awareness, Intentionality, Intelligence. AI only needs the last three to be useful and the last four to be dangerous. The AI does not need a theory of mind to collect enough data to be able to produce an accurate distribution of your actions for all interesting situations and then use it to arrange your environment so it can exploit you in line with its intentions.
>If you don't know how the human brain plays chess, it certainly appears intelligent. Once you do understand it becomes purely mechanical.
It is after all bound by the laws of physics.
basically i read it as: "hackers laugh at the pace of big corporations, but big corporations will prevail now that we need (gasp) 400k programs"
seems like the journalist started right, and then just wrote whatever the PR guy at IBM told him to write.
On a side note, I find it interesting that the author seems to equate artificial intelligence and e-mail integration as being approximately the same order of difficulty.