You're right it isn't. But also he's right, that the odds are against all startups - so a competitor also has the odds against them too. In an unproven market this is so much more true!
In regards to comments about IP :- compression and encryption are two obvious IP led areas. But also those working in the cutting 'technical' edge - such as high speed trading may well have solid IP.
The thing is a lot of SV based startups (a lot, not all!) are driven to solve customer/logistic problems not technical problems so they are less likely to have IP - unless (grrrr...) you count business practises as IP.
However startups out of a University incubator are more likely to be solving technical challenges and therefore have richer IP. As seen by the comments :-)
The light is shone so strongly on public facing SV startups it's easy for us to forget how many different types of startups there are and how they execute differently. For example most customer facing startups don't need stealth mode, they need MVP. Not true if your product is IP lead, in which case you need to be stealth while you secure IP rights and test the product internally.
There's a lot of massively useful advice coming out of SV - proven by startup after startup - but there is also a tendency to view all startups the same. The very fact that they are different means you need to have smart folk at the helm to navigate their own unique course based on the experiences of others.
It's a directly drawn from the premise that each startup is likely to fail. Stealing an idea doesn't make a company more likely to execute well and having an idea doesn't make a company less likely to.
Execution can make a difference on how likely you are to fail. There are so many examples out there where one execution won over even if the original idea was the same.
