Was pretty excited to see a nice welding of human data with AI, accessible via a “somewhat natural language” programming language, in Wolfram Alpha (https://www.wolframalpha.com/).
I enjoy asking it tough questions, but am not always totally pleased with the results. Since it’s a fixed database of knowledge, there’s not a lot of point in asking it about current events or other things you might think to ask a regular search engine.
For example, a question like “What is the name of Donald Trump’s exact shade of hair color?”, where the correct answer might be “burning shrubbery”, instead comes back from Wolfram Alpha with basically a bio of the aforementioned induhvidual.
Still, I would like to get into the Wolfram language a bit more and figure out how to drive what is present in the knowledge base, which is still quite massive. And one can even access some really cool functionality just by trial and error, by asking questions like: “how many golf balls fit in a gallon jug?”. It comes up with an estimate of 55. I don’t even care if that’s right, it’s just really cool.
By Chris Koeritz
(Originally published Oct 2 2016)