Ask PG: What would your next startup be?
PG, have you ever considered doing another startup? Or, even if you aren't considering it anymore, did you ever consider it after finishing up at Yahoo? If so, what would it be? Merry Christmas
If I had to start another startup in the classic sense, I might do RFS 1: http://ycombinator.com/rfs1.html
I did consider trying another startup once after Yahoo. I hired 3 hackers during the summer of 2000 to work on a prototype of a hosted application platform. I couldn't convince Rtm or Trevor to work on it with me, though, and I wasn't into the project enough to do it without cofounders.
This is an incredibly interesting problem I've thought a lot about, so I'll say the following despite the danger of turning this into discussion about rfs1. Since I first read it, I found myself disagreeing with the following statement:
We think they will mostly die, because we think we know what will replace them, and it is too far from their current model for them to reach it in time.
I think one of the fundamental questions is what type of content is valuable today (compared to what type of content was valuable two hundred years ago). Back when the newspaper business started, news were extremely desirable and difficult to get (and therefore valuable). Someone actually had to travel to the front lines, record the news, and then break it faster than everyone else in order to make money. The cultural artifact of this remains today - newspapers are still called newspapers, and when you turn on CNN you often see the term breaking news (despite the fact that every other channel is "breaking" the news as well). But this model has almost no differentiating value today - real-time publishing technology is so cheap that anyone can break news, and therefore no one wants to pay for it. News is a cheap commodity, not unlike orange juice.
Today we have an opposite problem - there is an exponential explosion of content, and people are willing to pay a lot of money for services that filter rare, excellent, original content from the rest of the noise (I don't count digg and reddit because they do a very poor job at finding excellent content). People are willing to pay for excellent writers, balanced and insightful opinion pieces, and articles about current events that actually teach us something, rather than tell us what happened. Some "traditional" publications have already figured this out - Economist, New York Times, and Wall Street Journal (although the future of the latter is in question precisely because of potential for declining quality). They didn't completely embrace this model and don't take nearly full advantage of information technology, but they're on the right track - they accept great guest authors, provide valuable and insightful opinion pieces, and give us articles that teach us things. In exchange, their customers have consistently payed up while other publishers have seen a steady loss of readership.
The trick here is branding - Economist has a world class brand that's recognized for their valuable content. Once they get around to using modern technology to take this model to its logical extreme, there is no stopping them. I suppose a startup could do it first if it figures out a good way to quickly build a brand (which shouldn't be hard, considering how little good content is out there and how many people complain about it), but fundamentally there is nothing stopping Economist from doing this - they're already in a perfect position to do it. All they need a little bit of technology (this should be easy), and a bit of a psychological shift in their culture. They're already on the right track - the solution is definitely not too far from their current model.
Also see Arrington on the New New York Times http://www.techcrunch.com/2009/07/30/what-if-the-new-new-yor....
Interestingly, old publishing technology (delivering a mass-produced hard copy into people's hands with no way to electronically communicate a typeset soft copy) essentially acted as an anti-trust barrier for newspaper publishers - having a national newspaper monopoly was not feasible (unless a magnate would buy up every local newspaper in the country). Today there is no such limitation - there's no reason for multiple newspaper publishers to exist, in the same way there is no reason for multiple video streaming sites to exist - YouTube can take it all. I don't think we can call this a solved problem until there's a YouTube for news content.
I can't wait for a startup (or for one of the existing publishers) to establish an infrastructure compelling enough for all of the best writers from all newspapers to switch to it (in addition to a huge number of really good guest authors who aren't professional journalists). It drives me crazy that the NYT is sitting on an excellent brand and a perfect position to do it, and they're basically doing nothing about it. They hire great journalists, and provide a medium for great guest authors, but they just won't take the leap and take this to it's logical extreme. When this happens, the poor excuse for good journalism we call "news" can finally die off and we'll be left to consume really good, valuable content.
I guess I'm drifting a little more off topic, but oh well.
Do you think there is potential for harm in a model where straight news isn't presented? One of the advantages of instant news is that there is little time for someone to weave their own views into what they write[1]. Despite how much I enjoy The Economist, its impossible to call their writing unbiased.
[1] I suppose its possible to take the opposite view on this. A responsible journalist will try to keep their views out of the article, but that is difficult and takes time, so their instant writing will be full of personal views. I guess it depends on how opinionated the journalist is on the news they are reporting and how unbiased they want to be.
I don't think straight news will disappear, to the contrary, I think the volume and level of detail of straight news will increase exponentially. It will just move to Twitter (or a similar platform). Build a good UI for Twitter trending topics, and you've got straight news. But I do think there will be little money for straight news content providers, since anyone can post a tweet - all the money in that space will just go to the publishing platform (looks like it's going to be Twitter).
The reasons why news and journalism are tightly coupled are completely rooted in old technological limitations - newspapers were the only ones who could deliver breaking news, and they were also the only ones who could publish opinions. So journalism is traditionally perceived as a mix of the two, to the point where the idea the two can be separate is alien to people. But once we take modern communication technology to its logical conclusion, it looks like news and journalism will go their separate ways - there is no longer a reason for them to be tightly coupled. Twitter will handle the news delivery, and some other platform will handle original commentary delivery.
So no, I don't think there is any danger at all.
1) Investigate journalism. In this case, someone is digging for facts and looking for connections. Tweeting, or any other instant platform, may not be the right medium for this. This looks like one area where news and journalism would need to remain intertwined.
2) Reporting from dangerous areas or areas without good wireless coverage. If someone is reporting facts from Darfur, it seems unlikely that any instant publishing platform could work.
Still, it appears you are probably right for a large number of situations. It looks like the situations I'm describing would need to stay closer to the current model for some time.
I think what you're digging at here is that you can't just provide a platform for freelance writers to write their pieces and pay them according to popularity - this model doesn't cover all cases. If there is demand for investigative journalism (which often, but not always, requires funding and cooperation) and reporting from dangerous areas, organizations would have to form that fund these endeavors. But, no matter how profitable this might be, this isn't a scalable business, so it isn't lucrative from a startup's point of view.
It sounds like organizations like Associated Press will eventually transition to handle scenario two, which they'll publish via Twitter (once it lets content providers monetize their content), and some newspapers will transition to scenario one (once one top-brand newspaper or a startup builds a value-added publishing platform). The rest of the newspapers will simply die and many of the good writers will transition to a freelance model.
Of course this gets very fuzzy very quickly, so most of these predictions will likely not pan out this way, but it seems pretty clear where the overall trend is going - technology will decouple a few concepts that are tightly coupled today. News would be decoupled from high quality written content, and production of content will be decoupled from its publishing and delivery.
I don't see much new in the world now vs 15 years ago that would shift the economics to high quality publishing. Technology and the internet can provide a better platform for crowdsourcing journalism or easier dissemination and aggregation of information, but these are different from just creating and consuming high quality content.
I think what's new is the ability to personalize the information. One way to think about it is that industrial revolution was all about economical mass production, and one of the first things to be mass produced was written text with the help of the Gutenberg press. No one knows what information revolution is really about yet, but one of the promises is economical mass customization, and the news is a perfect first candidate for this (more distant candidates being economically producing clothes that fit you perfectly, for example).
In a related area, Google seems to think they have mostly exhausted the opportunities to improve their search results in the same way for everyone: they have employed IP geolocation for ages and now they even track your search history by default in order to show you different results from what everyone else is seeing.
It seems to me that all the content aggregators and time-sensitive information availability via Twitter made news a commodity - there is nothing differentiating a Yahoo news article from a CNN article from an MSNBC article, from Twitter's trending topics feed. So how can a paper differentiate itself but to provide quality content?
I think what's new is the ability to personalize the information.
I agree, but I think it's only one piece of the puzzle. A "YouTube for value-added written content" would effectively be a publishing platform that connects freelance writers, content production organizations (like newspapers), and guest authors to their audience. Such a platform would have far more content than any given person could possibly consume, so it would need a personalization system (based on search, a data mining recommendation engine, etc.) But before you get there you have to build a compelling enough platform for good writers to switch to. There's a bit of a chicken and the egg problem here, so it would require a lot of seeding work to be done.
Of course this is just my view of it. It seems pretty clear to me, but there have been lots of things that seemed pretty clear that turned out to be completely wrong :)
15 years ago you could say the same thing about different newspapers instead of sites.
The publishing platform you describe is a compelling vision, but it is important to understand the differences from what we have now: e.g. Blogger.com, or the web as a whole as a publishing platform with aggregators like Digg/Reddit/HN trying to single out the high quality articles from that. If writing for your platform is important, it must provide some tangible benefit, like a way to achieve consistent style or quality or dealing with some metadata not available on the web.
Of course everything I say may be nonsense as I am trying to understand the problem space myself.
That is true, but it didn't matter than because the publishing technology for newspapers limits them to a small geographic area. So old-style newspapers only compete (and have to differentiate themselves with) other newspapers in the area, not all newspapers in the country. This is what kept them going. It's no longer true today, which is what made news a commodity.
People want to be sheltered from news which goes against their beliefs. Currently, the segregation of news sources does an adequate job of this. Right wingers watch fox and never need to learn about evolution, left wingers read the NYT and never need to learn about climategate. See also the reddit downmodsquad.
But this filtering scheme is pretty weak. I think the next stage in news is pure entertainment and ideological reinforcement. Ignoring ethics, this is what most people want.
(The only exception I can see is people who might be punished by the market for incorrect facts and the intellectually honest. But such people are a small minority.)
Most people don't hold extreme points of view on any subject though, it's just that zealots are most vocal.
Stereotype much? (BTW - There are several "right"s in the US, and only one of them, a fairly small one at that, disagrees with evolution. Others do think that teaching preschoolers how to fist is objectionable, but it's not clear that fisting has much to do with evolution.)
It turns out that Limbaugh etc often cite NYT articles. I'd guess that a reasonable number of folks actually followup on such cites.
The reverse - not so much. Otherwise "good people" wouldn't assume that "the right" is as cocooned as they are.
Would you only consider pitches that involved applying at YC? Or would you consider getting involved in startups that are going a different route?
I believe what you're saying is that any startup effort should be at least six months long (or you're doing it wrong)
Is that a true statement in every case? I've been reading a heckuva lot of startup information over the years here and I'm not so sure it is.
Maybe I missed out on something somewhere. Care to enlighten me with a link or an article or something?
I was also trying to regurgitate what I saw as the summed wisdom of the writings of PG, namely http://www.paulgraham.com/die.html and the advice to iterate quickly.
I suppose in many cases it is possible and desirable to prove your startup completely infeasible during the time between YC Application rounds. It's better to figure out something won't work than to go with something doomed to failure. But it's even better to iterate that initial idea to something that people actually want. I do think that if you are trying to tackle news/journalism that 6 months is not a long time at all to declare failure; there are so many different possibilities for applications and companies.
This is my excuse for not starting a startup. Startups are stressful. Why do it if you don't need the money? For every "serial entrepreneur," there are probably twenty sane ones who think "Start another company? Are you crazy?"" -- http://www.paulgraham.com/notnot.html
I'd think that he'd be working on one of these if were willing to spend the time.