Boy I hope that's not what people really think.
Ideas are everything. Execution is everything. You get one wrong, you get everything wrong.
You have to think holistically top to bottom. I like his overall point, which is basically that ideas about execution are the ones that matter. But the ideas at the beginning matter too.
That's pretty much my take on the whole "ideas vs execution" debate, with a nuanced caveat or two. Neither a good idea nor great execution is sufficient to build a great startup, and I'd argue that both are necessary depending on how you define "good idea".
IOW:
bad idea, poorly executed: FAIL
bad idea, well executed: FAIL
good idea, poorly executed: FAIL
good idea, well executed: MAYBE
I say it depends a bit on how you define "good idea" though, because I strongly believe that a "good idea" does not (necessarily) need to be something earth-shattering, brand new, never-seen-before, etc. Sometimes your "good idea" can actually be something that already exists, and you plan to win by simply executing better than the competition.
All of that said, I agree very much with gweinberg on the point about ideas within the context of execution mattering. In fact, reading that just sparked a couple of actionable ideas on my part. Those are the articles I love reading on HN.. the ones where I walk away with a new TODO (other than "read this later").
In our community, "bad idea, well executed" is also a maybe - not a "fail" - on succeeding - as by definition if you're executing well you know when to pivot and aren't afraid to do so, or you are simply succeeding despite a bad idea[1], or are doing what it takes to push it through.
Also I'd like to point out that the definition of executing well is almost begging the question. To answer the question "Was it well-executed?", we apply the heuristic "Did they succeed? get numbers up? generate sales, investment, etc..". Then we are only 'testing' our own definition and every successful company will be "well-executed" and vice versa.
I'm not sure how to improve these definitions, but it's important to realize that asking how well a team is executing is problematic.
You would need a definition of "execution" that lets you point out teams that are executing well and failing, before you had a concept of 'executing well' that is useful and not just a proxy for 'succeeding.'
[1] Plenty of big, multibillion dollar product categories (McDonald's for one, two year mobile phone contracts for another), are a really bad idea. If either category didn't exist, I doubt an investor would think you could convince people to like it, on a national scale.
And, this is skipping the issue of refining the actual idea itself as you learn, which may not be quite as dramatic a shift as an actual "pivot".
Your point is definitely valid though, that there is a some semantic confusion around this whole discussion.
with good execution, bad ideas often lead to good ideas. I know more than one nontrivial business that essentially started out with a bad idea, but the execution was enough to see what was bad about it, and that there's a related idea that's good.
I've seen much fewer cases where, despite bad execution, a good idea shines through: sooner or later, usually sooner, the competition wins when the execution is bad.