Thursday 17 November 2016

google doodle or quick draw mcgraw

We learn of another novel mentorship opportunity of machine-learning that we can all interact with the form of the experiment called Quick Draw from the search engine’s labs—thanks to the Presurfer. It’s a challenging few rounds of play with assignments that aren’t quite the easiest concepts to limn with a mouse but you should really give it a try—with the Control Voice shouting out guesses as you sketch like celebrity contestants on that television game show Win, Lose or Draw.
It always strikes me as a little strange to consider how—when engaging in this sort of crowd-sourced science, that we might be ultimately outwitting ourselves, but then again, one wouldn’t withhold wisdom from a fellow human. It’s quite a dilemma that were stumbling into. Relatedly, I noticed recently that the image search of Google has become a bit more literate of late: often I use it to search PfRC for pictures I can vaguely recall but have no idea what I saved them under and found if I typed in a colour—even if that had nothing to do with the filename but remembered that the background was purple, the query would yield what I was looking for. Also, I noticed without nudging that on my mobile device I could set not a temporal but rather a spatial reminder—like a shopping list that would go off when one’s in the supermarket. I wonder if those features, those talents came about all without the intervention of programmers and were the fruits of artificial intelligence. When that does happen, would we even realise it?