I am getting tired of the cutesy easter eggs on Alpha. I suppose it's good for marketing, but every time I see one it reminds me of how limited expert systems are and makes me think how shallow their lexical analysis is. 'What is skynet?', like many other things, produces only an error.
I agree. They seem to have spent a lot of time entering silly Easter-eggs. Perhaps they should have invested that time in loading more data into the thing, so that it can actually return useful answers to my queries.
I still don't get why queries like "yearly income vs age" can't find any useful data.
They really need a query builder and refine search tool. It's hard to browse the data to see what's there; every time I search I feel like I'm throwing dart while blindfolded.
Additionally, when I mistype things like 'us pgdp / population' it comes back and asks if I meant 'US population'. What I want is a little form saying 'hmmm... US [pgdp?] / US population ?' where [pdgp?] is an edit box where I can correct my typo.
I approach fixing broken things by beginning with the assumptions that they are designed to work, and that examining each subsystem for damage will get me there. In the example above, 'US', '/' and 'population' are working query terms, so why doesn't W|A keep those current while highlighting the need to repair the meaningless 'pgdp'?
