Would be great if the next Google/Facebook/Microsoft was European and HQ'd in Europe, but it's hard to see that.
ML anecdotes - Schmidhuber makes a point of how LSTM was developed on European taxpayer's dime. Backpropagation was invented by Esko Linnainmaa (a Finnish researcher) in the 70s.
Universities or startups can't compete with finance/Google/FB level compensation, so that's where a lot of the bright minds go.
I also used to think that, and in the long-term I agree, and I have (i) done a startup in Europe (failed for lacking of funding) (ii) worked for a European startup (Prezi). But right now I work for Facebook (in London), and this is a good thing. You can go to these places and learn a lot, a lot more than what you can figure out on your own with your friends sitting in a garage or at one of the moderately successful European startups (moderately successful=0.1% of Facebook, that's already like $300-500M), which any sane entrepreneur here would be super happy over, fully so.
By "a lot" I mean cultural things, vision, strategy, how to set your own expectations, how decisions are made, what tools are available, etc.
I agree it's a good learning opportunity for individuals, but I still hope Europe will be able to produce new large, successful tech companies in the future. Right now a lot of the value is captured by US companies
For that the European home market needs to become easier to compete on, though, and it currently looks like it’s moving towards getting a lot worse instead.
ML anecdotes - Schmidhuber makes a point of how LSTM was developed on European taxpayer's dime. Backpropagation was invented by Esko Linnainmaa (a Finnish researcher) in the 70s.
Universities or startups can't compete with finance/Google/FB level compensation, so that's where a lot of the bright minds go.