Lenat's probably the best-positioned guy out there to back-infer some of alpha's design and limitations from a two-hour demo; this is the most informative summary I've seen.
I do hope it's everything it claims to be though.
In fact, reading this review I'm reminded of the article about Git and merge algorithms that was on the front page recently. In all likelihood, Alpha will be amazing at what it does, without anyone worrying whether what it does is what people want. Maybe people like Google because it's fast and reasonably accurate, and the human brain is still a really amazing filter once you narrow down the data set sufficiently.
You can see it when you search Google for things like:
"What time is it in Las Vegas" (notice that it knows, and it also knows that you probably mean Las Vegas NV but might mean NM, and that Las Vegas NM is smaller)
"MSFT" (e.g. a ticker)
"Who is the president of France"
"16 USD in GBP"
From the stories, it SOUNDS like Wolfram can answer a lot more of these kinds of queries than Google, which makes me think that the best way to use a technology like this is to sell it to Google to plug in as their pre-search pre-processor to replace the simple ones they've already developed.
Anyway, the point is, when it comes to being "trusted" there are a lot of software companies and products who Google doesn't really compare to.