> I read the article, but it feels like "don't build anything until you are sure users want X".
The article addresses that exact point:
> I think the answer's obvious. Because they designed the product backward. They didn't think "what do people want?", or even go Apple and think "what COULD people want, if we showed them why they wanted it?".
Your issue is largely addressed by their second Apple example. They're talking about the thinking that ultimately results in the product, and how Google's (and Sun's) thinking is focused on themselves and e.g. Apple's thinking is still user-orientated even if they have to "sell" an original concept.
So absolutely try something new, but as the article says, it still has to be designed for an actual audience rather than just being created to further that company's own interests without offering enough value for their market.