The points Steve makes in the early parts of the session were quite visionary, but when he got to the hardware parts, I guess it's quite easy to say in retrospe

The points Steve makes in the early parts of the session were quite visionary, but when he got to the hardware parts, I guess it's quite easy to say in retrospect what they got wrong. What's interesting, however, is that he does discuss this from several angles (my interpretation / memory below), namely that a) the hardware angle was a way to get people to use the software, b) a purely software venture couldn't sustain funding for the level of marketing required to break into the market, c) software seems to be moving slower than hardware (? - I guess that's the 80s / early 90s for you) and d) there's still a juicy hardware business that would be left off the table, if it were a software-only venture. So, it was a fairly well researched and opinionated bet, that just didn't pay off. I guess by tying the success of their software to their hardware, they ended up sealing its fate.