[gcanyon]

The problem WA is attempting to address is nearly impossible: to trust WA as a reliable source of information, you have to be confident it will be able to answer the question you're asking. If you work in a specific problem space, you can probably know that, but even if WA does know your particular area, you likely know even better ways to answer your questions.

Putting it another way, it's too hard to know what WA knows and doesn't know. I alluded to this in a post I wrote back when WA first came out: https://gcanyon.wordpress.com/2009/06/07/bing-wolfram-alpha-... "As Alpha grows and adds new problem domains it will become more and more useful, but it will continue to be necessary to understand what it can and can’t do, and how to get it to divulge what it knows."