I've been reading less of ycombo news lately since it was just mostly articles on how to get VC funding and recent hoopla, and not really interesting things on the edges/fringes of markets, society, and technology. Imo, those are what you have to pay attention to, in order to have any type of gauge as to the probable direction of the future.
I've submitted a share of "interesting" things, but they're usually drowned out by "easy reading" Hopefully, the human editor factor will allow the proposed indirect control of the swarm.
I've had some interesting discussions over at octonews--but it's a different niche--hardware/sci/tech/health. I'd love it if yc.news is able to turn itself into hacker news. I had missed the mid 2006 reddit, and hearing great insight is often more valuable than the actual article itself. Not that how to fill out a term sheet, analysis on facebook, or top ten things to get started on your startup isn't interesting, but one can only take so much of that in the last 3 months.
I probably wasn't the only one who wasn't too sure about refocusing the site. One negative is that for those of us outside the Silicon Valley or YCombinator, it's possible that the articles and comments would focus more on programming than startup advice. On the other hand, maybe it would be great to refocus every 6 months. It would be a very unique idea. This is not to disrespect what you've done--thank you for the articles, startup school, this site, community, and ycombinator.
I like new name, but honestly dont think the content itself will change very much. I've noticed most submissions are generally hacker-centric anyway, although having the name of the site reflect this probably more accurately hones the focus. I think generally the content being submitted will be a reflection of the type of users using the site, regardless of the official mission statement. Manually moderating only goes so far...
Considering that, I would personally like to see this site make a conscious effort to remain "grass roots" rather than focus on growth. I love this forum because of the relatively high concentration of smart people who post here. I doest look like PG has any monetary objectives with this site (hence no ads), so I think simple word of mouth between hackers rather than deliberate promotion will keep content quality higher going forward.
The proposed rating system is brilliant.. I give props to whoever thought of that one!
Edit: I just realized that the new weighted voting system could have an unintended chilling effect on voting.. If smart users worry that placing a vote might adversely affect their "vote weight" on a post deemed by the admins as a dumb story, they might just refrain from doing so all together.
I agree with the edit. A good system should be working smarter and more pointedly. If you are already implementing a system that is smart enough to detect if votes agree with a editors system, wouldn't it be better to just detect if everyone else's votes agree with your own personal ones and then show such things?
The whole notion of having an ivory tower that tries to shut itself off from other people because they degrade conversation is like security through obscurity. What this and reddit needs are a proper recommendation system, which could be based on tagging. Reddit says they are going to do this, but even then I don't imagine it will be an overly complicated tagging regime (even though they are taking long enough to roll it out). Once you get into basic tagging (eg just a number of strings) you really have a subset of an ontology, so the next logical step is to build it into a big proper one like freebase is doing, which has proper datatypes for each tag value, which can then be other objects in a database. In this way articles can have a set of tags, like a category, a rating of intellectualness, etc. These tags should be modifiable and addable. People should be able to have different values that reflect what they think of them. The tags that are most relevant should be displayed. An article could have a quality value that some could rate highly, and some poorly. Then the system, based on your own past tagging, can predict what default value yours should have. It's much like the netflix challenge (which is 80% solved), only its not just for one value of one tag, and the values are more complicated.
Anyway, I've thought this through too much because I am half way to solving it myself (hopefully!). We'll see if it works when I start pulling people away from here :p
I look forward to seeing where this idea leads the site, but I do have one suggestion...
While I'm interested in all things hackerly, I first started visiting this site for tips and stories specifically about startups. My point being: while most people here are all for changing the focus to anything hacker-related, I'm sure there are many newcomers who are looking specifically for startup-related info. Perhaps we can retain a sub-site, startup tag or filter that will show only startup news for those people?
I'm sure there will too, but a feature that allows you to only view startup news might be a welcome addition for the portion of news.yc visitors (which I'm guessing is a majority) that come specifically for startup information. At the very least it will differentiate the site from the likes of the MAKE blog, hackszine, etc, which I'm sure will carry a lot of the same content.
programming.reddit is occasionally pretty good, but sometimes there's just something interesting which isn't strictly programming per se. This fills that gap perfectly.
also: "Most forums degrade over time, but we don't think that's inevitable."
it isn't, it just requires eternal vigilance on the part of moderators/editors. the SomethingAwful forums are in the 10s of thousands and they're still quality because of Lowtax's willingness to ban people.
Doesn't seem to be a very thoroughly thought-out move to me. It is inevitable if current users are willing to compromise. And given the amount of gray-area, I think they are.
Quality doesn't just vanish. It's eroded away like the hair cells in your ear. When you realize it's gone it's not going to come back without clever engineering.
So it's not just as simple as changing the name of a forum, declaring a shift in the theme, that would move the forum in one direction or another. Unless you have superusers and/or admins. The analogy is to slashdot, or to reddit. Keeping the reddit-style system means increasing popularity leads, without a doubt, scattered focus. Keeping the focus implies a shift in the current moderation system (unless there are invisible superusers lurking around).
The flaw of forums is not in the rules, not the theme, not the users, but in the system design. Change it, or compromise. Either way, whatever is written so confidently on hackernews.html will not be as simple as "I decide it so."
The exception to my argument is if growth of userbase isn't a goal. I'm not sure if that's the case here though.
My biggest worry is a slippery slope. For instance, lately I've been reading lots about economics. A lot of it is stuff I would consider hacker-worthy. However, it might not interest some people, and others might have philosophical objections. Or even political objections. We're already, at that point, on our way downhill.
Paul, it would be good, if at this point, you make the Hacker News Karma rules transparent. It would be good to have a page that explains how the Karma works and the other rules as well. It was one of the most irritating things with reddit, and the irritation only increased when they got sold ;-)
Thanks for putting in the time and effort to express the site's intention. My experience on YC News has been extremely satisfiing re tech and creativity discussions.
Why is YCNews trying to be programming.reddit? I think there's plenty of "Hacker News" websites out there and there's no investor/entrepreneurship/startup news sites. The focus of the site will be lost now :(
So you're trying to be the next reddit.com since that's mostly what "hackers" are interested in?
Also, the choice of the word "hacker" is horrible. Yes, I know the difference between cracker and hacker but you ask anyone out there what a hacker is and they'll tell you it's someone who "hacks websites and steals credit cards." To change that perception is a next to impossible task. So, I'm sure we'll enjoy a bunch of newbies coming here and asking how to "haxor" their friend's email account.
I don't like this change one bit. Are you trying to increase the amount of traffic on this site?
As for the word "hacker," there is no uncertainty about what it means among hackers themselves. And our sense of the word is spreading, not contracting. I saw a headline in the NYT a couple weeks ago that used "hack" in the good sense.
Previous reddit lasted 3 months tops before it degraded into a lot of nonsense. What makes you think that n.yc will last even that long before it gets flooded with crap?
The whole appeal of this site was that you could spend 10 min reading top few stories and you'd get a pretty good idea what's going on in the YC, VC, angel, investment community. Now we'll have to spend tens of minutes weeding through all kinds of topic that just waste your time.
Why not start a new website with this new theme? Since YC i a seed stage investment firm, what exactly does "Hacker News" have to do with YC focus and mantra focus?! It is a subdomain of an investment firm so the focus and direction of n.yc really makes very little sense.
Reddit held out longer than three months. I remember remarking to spez during Startup School '06 that I had noticed a recent, slight downtrend in quality. That was around the time that random photographs started appearing on the front page, and a couple weeks after the famous "Paul Graham Eats Breakfast" post. But it was far from nonsense then: the majority of front page content was still excellent. Reddit launched in late summer '05, so that's at least 8 months that it held out.
I don't know how it will turn out, but I think reddit worked out really well. It's no longer a hacker news site, but that's not really a bad thing. It's a great service for a lot of people. Honestly, the extistence of reddit and programming.reddit means a lot of people already have fine places to be.
news.yc and reddit are both young, even in internet time. New members are important, and any forum must adapt to stay relevant.
This is good, I think. If it avoids the problems that Reddit ran into it's wicked cool.
However, doesn't this go away from the ideas of the social network, user-driven, "wisdom of crowds"? Do you have the problem now of the group not choosing which stories are good vs. which are agreeable to the human editors?
Yes from the write up it seems to be moving from "wisdom of crowds" to "wisdom of AI". The editors will be "training" the system to recognize a good story. PG is probably hoping that one day the system will be able to tell a good story from a bad one without any help from editors.
I don't think you mentioned the number one reason why you did this... to reach out to programmers in general and then trick them into wanting to do a startup via peer pressure, so we have more founders and you have more options to invest in. ;)
Awesome move. One thing I always missed on early reddit and Startup News was Slashdot's comment-tagging system; comments are not only voted up and down but tagged with "interesting", "funny", "insightful", etc. Also, there's an absolute score on Slashdot comments (1 -> 5) so it's easy to see off-hand what the really good comments are.
In general, my point is that on social news sites, the comments are often just as interesting as the stories themselves. This is why (at least I think) Slashdot is so great. More software and techniques to organize and enhance comment browsing would be really cool.
You are absolutely right about hackers caring about other things than business, but we already have a hundred places to get all that information online. The reason I liked this site is that it had a purpose: to provide news about startups. With a focus like that, the site is unlikely to ever go the route of reddit (the internet version of chain letters and "my dad can beat up your dad").
PLEASE PLEASE PLEASE change it back. If you want to launch another site for hacker news, I'll be the first to subscribe, but PLEASE give us back startup news!
Well, we'll either be seeing floating-point karma, or we just won't see story and comment karma increase anymore when we upvote them if we have a < 1 vote weight.
Ah, good point. We can just look and see what happens when we upvote a story. Unless, of course, voting power varies based on the popularity of a story -- then it's more complicated.
Something about the idea of oracles makes me recoil. The obvious problem is that this approach may reinforce groupthink, both algorithmically and socially. And groupthink is already a considerable problem in these sort of systems. But then... this solution almost represents an acceptance of the inevitability of groupthink, kind of an attempt to steer the groupthink. Hmm. Perhaps this method would work best by working specifically with edge-cases - the best and worst posts, gently massaging out the toxic before it gets a chance to grow roots. It may not even be necessary to include good posts, since the actual problem to be solved is keeping out the crap.
But I am excited to see this kind of community flourish again, whatever shape it takes!
I'm also wincing slightly at the thought of oracles. Could it not be sufficient just to clearly define scope and quality standards and let the current karma system do its thing? Regardless, I'm loving the new direction.
I can't doubt the motives. But I have a theory on why some days there isn't much to read and why reddit had great links in 2006.
At that time people put the best of their bookmarks on reddit. Later there wasn't much good content left in the stash.
IMO link submissions are unlikely to change. No matter what you do.
I think one way we can have interesting dialog on the site is by asking really good starting questions. Most questions so far have been about asking for feedback. Which are interesting for the people asking for feedback and the people giving feedback, but not that great for the rest.
Love the idea. Very curious to see how it will play out.
One concern is that the trolls will upmod all stories that are already upmodded (for instance, automatically upmodding all comments and links posted by PG) in order to increase their vote's worth. One observes similar things on reddit already. It is much easier psychologically to jump on a trend than to do something truly interesting, but the truly interesting and innovative is rarely awarded immediately with recognition.
That's fixed by scaling the vote weight offset by how popular a story is. That is, the first voters on a new story are scaled more than the later voters. Voting on pg's comments and stories probably shouldn't affect vote weight, for the same reason he's not on the leaderboard.
Yes, you're right. I was thinking the same thing this afternoon as I was driving around.
However, one problem that continues to present itself is the cumulative nature of karma accretion. Only if there is a completely separate scaling mechanism which determines vote weight independent of karma could be this avoided.
Otherwise even if an upmodder gets less for voting up PG (or Nostra) stories, they still get something.
That isn't necessarily a bad thing, though. It means they're participating in YC news.
Besides, I bet the vote weight curve isn't linear.. It's probably logarithmic, so you'll see diminishing returns.
The real question to ask is, "Why would people want a large vote weight?" The only malicious reason is to get something meaningless to the frontpage really quickly. But that will still render them powerless immediately if it's flagged as nonsense.
Besides, at least one other person will still have to agree with you to get it to the front page. This sort of guard is to prevent your votes from getting things to the front page if you upvote nonsense.
Yep, the stories are already much more interesting. I just blew an hour when I only sat down to check my email. Great. Guess I'm going to have to call in hacker to work.
I actually started reading /. again a while back. The quality of the discussion there has gone up quite a bit ever since the kiddies and trolls defected to Digg and Reddit.
The breadth of people who read slashdot is still an advantage I have not seen duplicated elsewhere. There are many sites where programming stories will be comented on by programmers. There are not so many where particle physics stories are commented on by particle physicists.
Random thought: this shows the importance of choosing good names. If they had called it "startup.ycombinator.com" to start then they couldn't rename it to "hacker news" now. Same is true in programming.
You've just created a way for people to game the system. Presumably, you'll submit stories you think are appropriate. Upmodding stories you submit will increase the power of the modder's vote. Avoid revealing the other oracles.
Since everyone will have weighted voting power, all you will have to do anyway to game the system is upvote anything with a lot of votes. Steadily build up your karma with this bandwagon method, and then bam spring some ad-infested blog spam upon everyone.
This can be accounted for by giving more weight to those who vote for a story early in its accumulation of points. The biggest voting power boost is given to someone who gives a story its first point.
My guess is that these self interested folks would create many accounts, buy cred with mischievous modding, and make a distributed effort to upmod a particular story. They wouldn't do this to upmod any random story. But news.yc needs to get much bigger in order for this to be a real problem. This is a real problem on Digg.
I'm pretty sure separate accounts are linked via IP addresses. And the one person who will go to all the trouble to spoof his IP address consistently for each of his 10 accounts isn't worth worrying about.
We're small, and we're going to stay small for awhile (hopefully), so this really isn't worth worrying about.
She's deleted them recently. Google Reader still has them, so:
"Although criminals harm society, and slaves are seeking freedom, from the point of the criminal or slave (who is a criminal, de jure, for running away), the outcome is the same: very negative feedback.
You mention "the value pyramid". I don't see that: if you get caught and "massa" cuts your big toes off, you are back where you started, plus a whipping and minus your big toes.
You'll be gimping around for a long time after that.
The only good thing is that at that point, you can't be sold off to work in the rice. But that's about it; you are pretty screwed.
The only way it works is if you've got a chance of getting away. Then it could be worth it. As most slaves had no chance of getting away, they were nuts to do it."
There's a racist crack, and a very unpopular viewpoint, but as far as I know she's actually debating ideas, not just baiting people.
I've submitted a share of "interesting" things, but they're usually drowned out by "easy reading" Hopefully, the human editor factor will allow the proposed indirect control of the swarm.
I've had some interesting discussions over at octonews--but it's a different niche--hardware/sci/tech/health. I'd love it if yc.news is able to turn itself into hacker news. I had missed the mid 2006 reddit, and hearing great insight is often more valuable than the actual article itself. Not that how to fill out a term sheet, analysis on facebook, or top ten things to get started on your startup isn't interesting, but one can only take so much of that in the last 3 months.
I cheer the new direction