... except that 2005-2013 or so was an AI winter. So it was too much ahead of the curve, also useless. Nobody was investing in AI at that time, not even Google (that was when they got started and they thought letting linguists manually write rules was "obviously" going to beat ML)

I mean except that I learned a lot and went into networking and internet which went well.