We (Virtualmin) got in and the problems we solve are big and take a huge volume of code (though we had a 9 year jump on the problem when we applied, in our Open Source projects).

Parakey got in, and their problem set sounded really big...but I didn't understand what they were talking about most of the time, so I may have misunderstood. They also had a leg up, by being Blake Fucking Ross and Joe Fucking Hewitt.

Versionate is a big project, but they were well-started by the time of application.

Octopart is a big project. They faked it until they made it, and did so beautifully (their presentation was among the best on Demo Day for WFP07).

Loopt is a big set of problems, most of them non-technical, and pg considers it the best thing out of YC to date (and I see why...Sam's kicking ass).

Zenter was hard. But they broke the rules and pretended like it was easy and actually produced something that mostly worked in a few months.

So, yeah, while the majority of YC companies are pretty small ideas that can be executed by a very small team, many success stories (or success stories in the making) out of YC have been big ideas.

is there a way, any way, for us plebeians to have a look at the presentations done on demo days?

I doubt we'd make them public, since (a) most of the startups haven't launched yet by that time and (b) even the ones that have tend to describe future plans they'd rather not have get into the hands of competitors.

One option would be to record them, then let the startups themselves decide whether to distribute the videos or not.

That's a good idea, though the startups could just repeat the presentation later in front of a camera, which might be a better idea (they can do multiple shoots, etc).

While it sounds like a good idea, that seems unlikely to work. Why would they put in the extra effort? I think it would be one of those good ideas that get postponed indefinitely.

Yes on all counts. I wouldn't want our Demo Day presentation disseminated because it kinda sucked. And I wouldn't do it again, because when I next present I expect a big fat check to follow within a month or so. But maybe I'm just lazy.

How can there not be any value in this? If not for the practice at least for the fanatical viewing this particular audience will do. And I'm also surprised at how little the companies communicate their product idea or vision. But maybe I'm focusing on the wrong thing.

I tend to agree. People may be underestimating the benefit that the increased awareness would give. Serendipity is powerful because "random" people often know the "right" people.

The two ideas aren't mutually exclusive. Some presenters might do better in front of an audience, some might find they can do better the second time around. You could mix footage to redo certain parts, etc.

We may start to do that, if only to show future YC startups what a good demo day presentation is like.

We like big ideas. Just say what the end goal is and what you currently expect to do first. There's always some core you can get done in 3 months, and we are now quite good at figuring out what it should be.

Sounds good, that's the answer we were hoping for.

What I'd like to accomplish: Become Master of Space and Time

What I expect to do first: Apply to YC

For some reason, I'm having some trouble with some of the steps in between. (grin)

Consider a tag cloud.

So 2006. What he needs is an iphone app with facebook plugin & twitter integration. Which is also a platform.

I was thinking what I needed was a sock puppet, or maybe a website with a name with too many vowels, like ooweblioo. I'm also considering a folksonomy of avatars, or maybe a paradigm-shifting web 3.0 application that virtualizes social dynamics in a new dimension.

Dang! That sounds great! I think I'll call the puppet Bob. Man, I'm almost there.

Wow, it IS like 1999 all over again, isn't it?

I remember shopping for capital in '99, and we had quite a few problems getting VCs interested because we weren't asking for enough money, had a CEO who actually had 20 years C-level experience and a name that didn't start with the letter e.

About 2 years after we shuttered and sold off most of the code to various people, EMC implemented something very similar and made a few hundred million dollars off of it :/

There's some similarity. Only this time, the big companies are buying up the little fish early in the game and absorbing them into operations -- which is probably a much saner thing to do than everybody betting they have the next Google.

We're seeing the beginning of "retail startups", where we can spit them out like six packs of Coke, fund them, get them bought up by Amazon or whomever. The industry really wanted to get to this point in '99. They've made a lot of progress since then.

"I like the way you think kid... here is 10 million dollars!"

"By the way, can we make it 4.0? 3.0 is so 1999"

-Anonymous VC

Tag cloud seconded. Then progress to vertical reflection of your graphics.

So you're trying to be a mashup of Sky in Google Earth and the Wayback machine? That's going to be tough...

(I'd vote for whoever is the first to do it and post it on news.yc!)

I think a good answer would be, "Anything that could be possibly scalated to a 'world domination' plan"