Ask HN: Pick startups for YC to fund
Hacker News users have many diverse perspectives on technology and business. Perhaps if HN picked startups, it would pick differently than YC. Maybe different startups would be motivated to apply, if they knew that the interviewing and deciding would be done by the HN community. Interesting things might happen, or they might not. We'd have to try it and see.
I ran this by Sam and he ran it by Kevin and we all got excited, so we're going to try this as an experiment. Starting today, there's a new track for YC applications: applying directly to the Hacker News community. We'll call it "Apply HN". Note that word experiment! We'll start small and figure it out as we go. But here are the initial conditions.
YC’s Fellowship program will fund a minimum of 2 startups selected by the HN community for this summer's F3 batch. (They name their batches sequentially.) The Fellowship is the YC program that fits best here since it’s designed to be experimental and inclusive and doesn’t require people to move to the Bay Area.
All the interviewing and evaluating will be done in regular HN threads, and everyone is welcome to participate. For this summer's batch, Apply HN will accept applications starting now and ending April 27.
If you'd like to apply to the HN community for YCF funding, simply post a submission whose title begins with "Apply HN" and explain what your startup does. Hopefully community members will ask you questions and good discussion will ensue.
If you'd like to help pick which startups to fund, simply jump into any Apply HN thread that interests you. Ask questions and post comments that you think will help the community make the best decisions. These will be regular threads, with all the same voting and so on, but with one additional rule: Be Nice.
Be Nice is a stronger version of our usual rule, Be Civil. Anybody who applies to HN in public this way is putting both themselves and their baby in a super vulnerable position. We're going to rise to the occasion by being not only civil, but nice. When interviewing startups, by all means be curious and probing—but only if you can also be nice. The word "nice" originally meant "not knowing". Then later it meant "precise and careful". And now it means "kind and thoughtful". Let's put all those qualities together.
At the end of the month, we'll rank the startups and YC will fund two. The ranking will depend both on upvotes and on the quality of discussion, similar to how the ranking of stories works. We can talk about this in the comments, but to answer one question I know will come up: Upvotes are an important factor but they're too brittle to rely on exclusively; doing so would encourage the wrong kind of trying to game the system. So we're going to gauge community interest both by upvotes and comments, and in case of doubt I'll make the final call—or better, figure out a way to put the final call to the community.
That's what we're thinking so far. If it seems unstructured, that's on purpose: we don't want to bias it along the lines of how YC already operates. We want to see what the community comes up with.
Questions or suggestions? Let's discuss and refine this together.
Edit: As discussed below, we'll add a top link for these a la /ask and /show. I won't get to that till later, though, so in the meantime, use the following link to find Apply HN discussions: https://hn.algolia.com/?query=%22apply%20hn%22&sort=byDate&d...
Edit 2: Here's the link describing how YCF works: A) There are many external factors that can implicitly cap the number of upvotes/comments. (time submitted, amount of competition, etc.) HN has repost rules to alleviate this problem: would Apply HN posts be able to repost too? B) Not to mention that it encourages sockpuppet voting/commenting, especially since there is a high reward for doing so. Product Hunt, for example, thrives on "how can I get exposure for my startup submission outside of the intrinsic quality of the startup itself?" and it would be an improper fit for HN. That's how the ranking of HN stories works now. It's a combination of upvotes and human curation by users (e.g. flagging) and moderators. So there's going to be some combination of technical and non-technical factors. How exactly will that work? We don't know yet. Let's see what happens and figure it out together. That's part of the experiment: we'll figure it out in the open and adapt to feedback. But let's wait until we have some data to look at. FWIW, my experience with HN is that genuine community interest is relatively easy to recognize. Another thing to watch would be the mean user karma of all votes.
A bunch of low-karma vote stuffing would actually bring a post down. If you can get a ton of massive karma users to upvote and game this, maybe you should just get in by default? High uncertainty is actually good. That's what makes experiments valuable - they reduce the uncertainty and thereby help us better understand something. In this case that "something" is the HN application process. So even if this process has potential problems, the experiment is still worth it. More interesting to me though with all this is what it say’s about the future of scaling YC. YC’s mission is to enable innovation, and funding more startups hopefully leads to more innovation. The full YC program, already spread across two demo days, is probably close to the limits of how many startups can be funded in one go. The Fellowship, by being virtual, is much more scalable but one of the key bottlenecks is always going to be reviewing applications – this experiment could help remove that bottleneck. (another bottleneck I see is office hours, but getting sci-fi for a moment maybe YC research can develop a chatbot for that one day!). I think the equalitarian thing to do would be to encourage apply hn's to be as self contained as possible to avoid biasing on the quality of the landing page or video or whatever else would profit from the exposure I think we'll actually learn more during this experiment with less rules on format up front. This problem happens to me so much I just ignore most Show HN's. I open a mainstream website to find who they are, what problems they solve, their solutions in a high-level way, and datasheets or whatever giving technical specifics. Many Show HN's here are one page of stuff that I read and read all the way to the bottom before I even begin to grasp what the hell they're doing. That could just be due to them working in a specialized niche with its jargon. Yet, I have an intuitive feeling that it's often a problem with the presentation because they're usually intending to reach (a) current, jaron-knowing users of specific tech and (b) future users who know the concepts or needs but not that specific tech & jargon yet. So, at least a few sentences or a paragraph for people in (b) might be helpful. And at the top to save time. Random sorting is a good idea. We'll think about that. The HN community might just be a better barometer, hope to see this experiment work out and bring attention to worthy and unnoticed startups. I've seen this done before and it's incredibly efficient. It’s easy to form some really bad habits when you sit in a position of power to judge the potential of a person, a team, an idea and their execution—believing that you know better and focusing your time on finding weakness. The best investors don’t spend a lot of time on what can go wrong. They already know the odds are against every startup that ever comes into existence. They already know every startup is a shit show. Those will be the reasons why all the other investors will miss out on an unpredictable opportunity. The best investors try to figure out what can go right. They dream a little with the startup and they then sell that vision back to the founders. Remember that the big wins in startups come from the margins. For you to find what no one else could have predicted, know that it will take the shape of something that isn’t obvious. Being nice gives you a good foundation for being open and optimistic, which is what we strive for when we read applications here at YC. Thanks again for trying this out with us. I'm really excited about what we discover together. Thing is we don't find out usually until after we invest what the real problems are...so we optimize for people we want to spend a lot of time with and work on these problems/issues whatever they may be. Most founders feel imposter syndrome after they're accepted. The great thing about being part of a batch is you realize you're not alone and then learn that shit show is actually a default state. When you learn that, you relax and focus on the resourcefulness you need to solve your problems systematically. It always seems to me that "grit" and "stubbornness" are similar, and being stubborn to the degree of being unable to give up a hopeless dream seems to be a problem for startup founders. How do you distinguish between these very similar traits when you interview someone? To use a silly example: A flower buying website might do best to avoid spending time/money crafting promotions aimed at single biker dudes who live life off the grid. Even though there is possibly large group of folks that are under represented in the rebel luddite flower purchaser demographics. My burn rate is close to 300-400/mo, less if I eat less per day. #grit Beyond that what else is there? When I was trying to learn to code, I was an undergrad with literally no money, but I had a prepaid college meal plan my parents had bought me. So I ate one meal a day for a month and sold off my extra meals at a discount for cash to buy coding books. For example, if your van runs, I can think of several possibilities, especially if you don't need it during the day (i.e. drive to a library or somewhere where you can work for free all day and then you've got a HUGE asset available to work out trades and deals all day every day). Be Nice is a good goal for Apply HN since it promotes civil discussion and stops the threads from devolving into semantic "lol no https u suck" that some Show HN threads tend to devolve into. Users are not nice. They don't comment on HN or Twitter, they just stop using it without fanfare. And what determine a startup's success in the real world is if people want to use it. Hindsight is 20/20, yes. But I do have concerns that "Be Nice" may dissuade commenters/potential customers from pointing out legitimate, immediate flaws. I don't think HN users are going to have trouble doing that, but there are different ways to do it, and the nicer ways have the critical property of opening the discussion further instead of shutting it down. We HN users are used to thinking of ourselves as the scrappy underdog, calling out incorrectness and badness with pristine honesty and with no particular effect on the situation, except occasionally a righteous one like forcing Google to do customer service. We don't think of ourselves as being in a position of power, but that's actually an illusion, as many who've walked into a wasps' nest of critical comments and come out stung can attest. To me one interesting social aspect of this experiment, quite separate from who gets funded, is that it unambiguously puts the HN community in a position of power—a perspective that we're not accustomed to, and which requires developing different habits (edit: I mean a higher standard of conduct). Maybe those habits will end up translating to the community as a whole (yay!) Or maybe they won't show up at all (boo!)—but at least we can all keep reminding each other to be nicer in these threads. A favorite expression of mine: "Sufficiently advanced political correctness is indistinguishable from sarcasm." I had always seen this, but never actually looked into what it did. For someone like me who typically makes a litany of edits immediately after posting, this is revelatory. Thanks! For whatever reason, the textbox just isn't as good of a preview when compared to viewing the actual post. I know. Good lord do I know. I have few in mind like Unicorn engine (http://www.unicorn-engine.org/). I would be supremely interested (as a victim of this really aggravating compulsion myself) in an A/B on your end, to see whether people do this less if the box is visually very close to the final product. My gut says yes, and it'd be useful data for lots of designers. (Anybody reading that's tried it?) Something unconscious about the flow of paragraphs, color, length of my lines, something drives me back to editing religiously until the software no longer lets me, on any site. It actually bothers me and I've already done it twice on this comment. (Make that five times.) Not only would that make you vulnerable to changes in the legal / enforcement landscape, but there's also a growing backlash in some quarters (including on HN) against this type of business. Can you provide any assurances that this business would be sustainable if you were forced to rely on 100% legal and taxed host properties?" I think the confusion is in whether conflict-avoidance is the same as nice-ness. I certainly was brought up to mostly think that way, but now that I've spent a lot of time arguing about math I have let go of some of that training. The former puts the applicant on the defensive whereas the latter encourages discussion. And, unfortunately, I have work to do and a headache, so I am not likely to be able to improve on either of their explanations. But you simply can't get there from here in terms of building something if all people do is tear it down. But I welcome this experiment if only in hopes that it sets a social precedent and teaches more people here how to communicate better. Because if you think being nice involves not pointing out legitimate flaws, you really have a lot to learn. It actually isn't nice at all to let people keep shooting themselves in the foot, but it is possible to point out the connection between aiming the gun down while cleaning it while loaded and the holes that keep showing up in their feet without saying "God, what a fucking idiot! Do I even need to TELL you how stupid that is???" Yeah, when it comes to ideas, you often do need to tell people that because metaphorically shooting yourself in the foot isn't anywhere near as obvious to most people as literally shooting yourself in the foot. 1. stick to "just the facts", eg, "lack of https support will expose your users to security vulnerabilities" 2. list potential solutions if possible Are we betting on the viability of a product or idea? Are we betting on the founders? Are we betting on the team's ability to deliver? So many questions... I can't answer any of them but I'd welcome people thinking bout loud and look forward to reading these ideas. (: Edit: recuse not revise What gets people tripped up is that you sometimes judge a founder indirectly by seeing how they communicate an idea, looking at their ability to deliver, and hearing about how they think about the market. The answers to those questions are important, but we put it in the context of what does it say about the founder. One advantage HN has is more time to think. HNers will look at a much smaller stack of applications, and won't be restricted by the brief interview format YC has. Could someone from YC please explain what's going on here? Perhaps I just missed it? This should be carved in stone :) If you think of HN users as 'investing' their time and energy into which stories they read and discuss here, then arguably this is kind of already happening. One way of looking at this experiment is to ask what happens if we take the existing system and permute its feedback loop, i.e. take one of the outputs (startups gets funded -> HN discusses them) and make it an input (HN discussion -> startups get funded). It's impossible to predict how a complex system will react and it's a fun kind of problem to think about. Of course, such permutations often just produce an unstable mess. That could happen here, i.e. the whole thing could turn into a trainwreck that produces no clear signal. I agree that people that didn't have HN accounts previously should definitely be allowed to participate both as submissions and in the comments :) , only that the upvotes part could be gamed. The upvote/commentary is likely going to be highly problematic: A) I'm pretty sure Dang can't list all 9 accounts I've used on HN over the past few years. I'm also pretty sure 9 accounts in good standing [1500+ karma] would be enough to seed an initial vote if handled carefully. The ones he'd be able to name are the 3 most recent. This being the case...yeah. The danger of sock puppetry is way too high, particularly when combined with strategic downvotes and VPNs. B) Anyone with control of an existing community/following can use it as a targeted campaign. C) The reward from bypassing these functions is substantially higher ($20,000 + professional advice from YC) than previously existed for YC. D) If you want to stick with a vote-based system, I'd offer a larger bounty [say, $50k for manipulating a post to the top of the pile for Apply HN] for finding a way to bypass the vote brigading and other countermeasures. This it is really the only countermeasure I can think of to counter manipulation problems. E) All the other external factors (time of submission, competition around that time period, repost rules, etc.) are likely to bias the process in unexpected ways but that is probably less problematic than direct manipulation. > I think we're on the same page to some extent. What I meant by 'ranking by comments' is that we want the ranking to depend on the quality of the discussion. That's the one thing that can't be gamed. Put a bounty on that to test that theory. I'm pretty sure someone could create a network of socket puppets that had "quality" discussions between them if the reward was $20k. I was just pointing out a direction someone could go. Remember, you're dealing with hackers here. I worked my YC application from every angle and tried to get every possible edge and advantage. Not necessarily because I wanted to game the system, but because I'm a hacker, and it's my nature [1]. On the whole, this is a really good idea. It's kind of mind blowing that YC would seriously try this. The concept itself is a brilliant hack on the idea of startup funding. Maybe they'll find the next AirBnb/Cruise/Dropbox, but probably not. Either way it's a great example of pushing the envelope to find new ideas, which sama has been all about lately. In the worst case, we'll at least get a lot more data for the voting ring detector. But I'd emphasize the qualitative side of the evaluation process -- it should be based on a somewhat opaque mix of upvotes, thoroughness of discussion, and thoughtfulness of answers. I'd almost treat the various 'Apply HN' threads as open source interviews. Upvoted and the like should be a ROUGH filter, but I'd hope that the YC partners will look at the answers in the threads as a crowdsourced interview and select on the merits more than anything else. Anyway, excited to see how this experiment plays out. Then the feedback and comments would be open for say another week. Putting time restraints would help give applicants approximately the same amount of attention. And also allow members take out time and ask all questions they want. In addition, can the submission be restricted too so we do not see 400 words submissions? I'll go for a maximum 2 mins video pitch alongside a link to a website. Finally, great idea! I would be wary however of learning very much with a one time experiment with only two funded companies. That would be similar to giving a prospective angel investor the advice to invest in two companies in the next thirty days and then no more. I think I have to disagree with you there. Traditionally, HN users are highly interested in the latter. Your other point is true of course, but we're using the word 'experiment' far more loosely than that. This is not science, it's just 'I wonder what will happen if we try this crazy thing'. Starting bigger doesn't seem prudent.299 comments
[ 3.4 ms ] story [ 270 ms ] thread
Maybe you should just implement tags already and let people subscribe to them or block them. That seems to be what this is kind of turning into anyway. Apply HN
Ask HN
Show HN
(unofficially) Tell HN