←back to thread

Tell HN: Winners of Apply HN for YC Fellowship 3

196 points kevin | 1 comments | 05 May 16 01:16 UTC | HN request time: 0s | source

Last month, we decided to reserve a few spots in the next Fellowship batch (F3) for the Hacker News community to decide who they’d like to fund. Startups applied publicly via HN and the community “interviewed” and voted for their favorites.

Context: https://news.ycombinator.com/item?id=11440627

We ran a poll for the top applications and the voting was so close that we decided to fund one extra startup. Here are the winners:

AutoMicroFarm (264 points): https://news.ycombinator.com/item?id=11454342

Feynman Nano (208 points): https://news.ycombinator.com/item?id=11443122

Casepad (200 points): https://news.ycombinator.com/item?id=11452884

I’ve talked to the founders of these three startups on the phone already and I’m really excited about working with all of them. We’ve disclosed all the vote totals in the original poll thread (https://news.ycombinator.com/item?id=11615639). Of course, the application that got the most votes isn’t on the final list and we’ll discuss that in the thread below.

We received 343 applications via Apply HN and over 1700 comments were generated across those posts. I was quite impressed by the quality and depth of the discussions on these applications and really loved the moments when HNers would take the time to provide quality feedback to the founders on their applications.

Thank you to everyone for participating in our little experiment. It takes a lot of bravery put your passion out there to be judged publicly and it takes a remarkable community to treat that courage with kindness and respect. It makes me very proud to be part of HN.

While we haven’t definitively decided whether we’ll do this again at this point (we’ll want to see how the companies do in the batch), I’m delighted and optimistic about what the community accomplished here.

We’ve already received a lot of great feedback from many of you on how to do this better, but please feel free to share more below.

Show context
cperciva ◴[05 May 16 02:26 UTC] No.11633569[source]▶
I've had a bit longer to think about this than the rest of you -- I discovered that the vote totals were being inadvertently leaked earlier today -- and I have to say that I'm impressed with the quality of the results. Leaving pinboard out of the picture[1][2], I think the voting selected a group of companies with tremendous potential: AutoMicroFarm has the potential to develop scalable technology which completely changes how food is produced, Feynman Nano has the potential to save hundreds of thousands of lives, and Casepad has the potential to make a huge industry dramatically more efficient[3].

I hope YC does something like this again in the future, but I'd suggest a different approach: Rather than asking HN to help select startups to fund, I'd suggest asking HN to help select startups to interview. This would solve the problem of needing an out-of-band mechanism to determine if an application is "real" or not; worst case, YC would end up paying travel expenses for some companies they decide not to fund. I suspect that the advantage of having extra eyeballs ensure that they don't overlook promising startups[4] would easily justify this -- not to mention the possibility of saving YC lots of time on in-house reviewing.

The one biggest danger I see with this is the potential for vote brigading; I suspect that we would have had more of that if it was announced at the start that votes would be a significant deciding factor. One possible way of solving this would be to limit voting choices, e.g., ask each person to pick one out of a small subset, so who only turn up because they want to support a particular candidate would usually end up filtering themselves out. I suggested this to Dan, but he thought that there wouldn't be enough voters to make this feasible; I'm not sure I agree with him.

Obviously I have no idea how this experiment is being viewed from YC's perspective -- and it sounds like YC won't know how they view it for a while yet either -- but as an external observer I'd say that this was a very interesting and very successful experiment.

[1] I'm sure that most people who voted for pinboard did not do so because they thought it would be a good investment. This gets back to the "it's not really clear what YC is trying to accomplish with YC Fellowships" problem which I've mentioned in earlier threads, but the original call was to "fund startups", not "fund your friends".

[2] I think that YC could gain a lot by creating some other mechanism to bring people like Maciej into the system -- something like an "honorary YC founder" status. But I don't think funding is the answer here.

[3] I'm a Canadian and not particularly familiar with the US legal system; but I've seen enough of it to know that (a) there's a huge amount of money there, and (b) they desperately need an infusion of technological competence.

[4] I'd be interested to know if any of the three had applied to YC via the normal route: Did HN identify good startups which YC missed, or would YC have funded these three anyway?

replies(5): >>11633611 #>>11633660 #>>11633740 #>>11633748 #>>11633777 #
jonwachob91 ◴[05 May 16 03:10 UTC] No.11633777[source]▶
> [4] I'd be interested to know if any of the three had applied to YC via the normal route: Did HN identify good startups which YC missed, or would YC have funded these three anyway?

During my phone call with Kevin this was one of the questions that came up, and the answer is yes, Feynman Nano applied to F3 through the traditional route. But we got buried under a lot of other applications.

As appreciative as I am to Kevin for selecting us, I'm even more appreciative of the HN community for giving us an opportunity.

replies(2): >>11633836 #>>11633958 #
cperciva ◴[05 May 16 03:26 UTC] No.11633836[source]▶
During my phone call with Kevin this was one of the questions that came up, and the answer is yes, Feynman Nano applied to F3 through the traditional route. But we got buried under a lot of other applications.

Based on this alone I'd say that Apply HN should count as a success: It allowed YC to catch a startup which fell through the cracks of their existing system.

replies(2): >>11634861 #>>11634975 #
1. danieltillett ◴[05 May 16 09:52 UTC] No.11634861{3}[source]▶
This might be the case, but I think we should wait to make a final decision if it is a good way to catch startups that fall through the cracks. We really need a random selection to know if HN (and YC) are adding alpha.