Bomb-Throwing Entrepreneur Peter Yared: Social is “So Over,” Machine Learning is Next Wave

Entrepreneur Peter Yared doesn’t mince words. In April, after TechCrunch misreported some of the circumstances around a Facebook employee’s termination, Yared wrote a widely read post titled “Why TechCrunch is Over” in which he called its founder, Michael Arrington, “insane,” adding that it “must be hard to live amidst a rapidly declining site.”

In more recent posts, Yared has called Twitter “primarily a broadcasting platform with very few active users” and unusable for “normal people.” He has also suggested that if he were to start a company today with either entrepreneurs Mark Pincus, Evan Williams, or Mark Zuckerberg, he’d go with Pincus “given what we now know” about Williams and Zuckerberg. (Both have been accused of elbowing their early cofounders out of the picture.)

It’s no surprise then that Yared — who has founded and sold four companies, including to Sun Microsystems, VMWare, and Webtrends — is very fun to chat up. I caught up with Yared earlier today. Our conversation has been edited for length.

You recently wrote at length about how Twitter could improve the customer experience, as well as make money. The company seemed to only half-kiddingly offer you a job afterward.

[Laughs.] Yeah, they were [very nice to me] for a while there.

Your advice was fairly comprehensive. What did you leave out, if anything?

That Twitter should pick up Myspace. Twitter is a publication platform for brands and media that has supplanted MySpace. Lady Gaga used to do everything on MySpace; now she does it all on Twitter. They could buy MySpace, shut it down, and migrate the accounts and all the page admin stuff – which is OpenSocial – and immediately double their traffic.

It’s not going to happen, though. I don’t think they have the gumption to make a huge strategic shift like that, something that could blow up in their faces. There’s also [the concern that people will think]: Is this a loser buying another loser? They’d kind of have to message around that.

Where do you find the time to blog? You’re still at Webtrends, plus you somehow started and sold your fourth company, Postano, right?

I started that as a side project, even before Flipboard came out, as a real-time newspaper type thing that pulled in the news and videos that friends had posted and so on. It was just for me, but I’m on the advisory board of TigerLogic [a data manager and application developer for business software], and I was like, ‘Look at this fun thing I just wrote.’ And they said, ‘Why don’t we buy this from you?’ I’d stopped using it because I realized that socially curated news is just boring. It’s just another feed. But then we got all these in-bound requests from brands that said, ‘We love how it looks; can we get one?”

The service basically pulls in Twitter feeds and Facebook status updates and YouTube videos, providing its customers with a more interactive way to tell their customers what people are saying about them across the Web.

Right. I’ve been running Facebook campaigns for four years, for everyone from Lady Gaga to State Farm, and they have teams of people updating [their Facebook campaigns] every day. Meanwhile, at [their separate, corporate] Websites, nothing goes on aside from the occasional press release that pops up. Now a brand can point to all their activity on social networks from their own site. They get their own real-time news site that they embed, instead of having a news page.

And they pay between $10 and $1,000 a month for the service. Will you see any of that money?

Yes. [TigerLogic] paid for all [Postano’s] development and I get a percentage of downstream revenue. I think it’s much better to find a corporate parent, have them fund [a technology] and get a percentage of revenue, instead of going and raising capital.

So the idea is that it doesn’t make sense to cede so much control to Facebook because it’s got its limitations.

Yes, social is so over. I’m so tired of it, my god. Hello, Mr. Brand, you’re putting all your stuff everywhere but your own place.

Facebook is a photo sharing site, an event site, and a contact list. It’s basically Flickr plus Evite plus Plaxo. That’s the primary use case for most people and that use case is already locked in people’s brains. The company is adding features, so that now if someone wants to see movie or get the news, they’ll [ostensibly] come to Facebook. And they keep trying to do these bizarre social ad units, but it [neither will work well].

I’ve come to this new theory that you can’t put the ad into social. You have to put the social into the ad. People don’t want to see ads in their [Facebook] news stream. So-and-so likes Starbucks? [Expletive] off. It would be better if Starbucks had a banner ad that included what your friends are saying about Starbucks, or that said, ‘Here, give this coupon to five friends.’ But keep it out of [expletive] stream.

Do you think then that Facebook is overvalued?

No, not at all. I said a long time ago that Facebook would be worth $200 billion. The companies that own their segment are worth $200 billion. Google owns search; it’s worth $200 billion. Oracle owns everything [relating to] the client server; it’s worth $200 billion. IBM owns mainframes; it’s worth $200 billion. Microsoft owns the desktop; it’s worth $200 billion – plus or minus. There’s no question that Facebook will be right there, too. There’s still plenty of money to be made and a lot of media spend allocated to it. Social is still a growing market.

So when you say social is over, you mostly mean…

That it’s way to too late to start a social company today. Seriously, who would do it but bottom feeders?

Any thoughts about how valuable location-based services like Foursquare’s can become, or do you consider them a subset of social, which Facebook already owns?

Foursquare? Who cares? Honestly, I’m the mayor of something – so what? It’s a coupon delivery system. I checked into a Starwood to get Starwood points recently because I’m a mileage whore. But it could have been Facebook Places.

What’s next then?

The one thing nobody is talking about is the Google car that drives itself, They built a car that drives itself. Think about that for a minute. It doesn’t get into accidents. It doesn’t run into people. Let’s say Google spent a million dollars per car [in development costs]. Well, it will probably cost $500,000 in a few years, and $3,000 a decade from now. That’s machine intelligence, and we’ll see it not only in self-driving cars but cheaper Wall Street trading systems and everywhere else.

This isn’t artificial intelligence. Google’s cars, for example, are just really good at pattern recognition. Some of us geeks disassemble the different pieces. There’s the red-light detection, the pedestrian detection. They must have a bunch of pattern recognition subsystems that report up to a cortex that makes a decision. It’s not like one big logic tree; it all has to be happening in parallel.

Ad optimization systems are also [good at pattern recognition]. There are a lot of computers doing a lot of work on behalf of humans. You don’t have humans saying every week, ‘Can you send me the analytics?’

Machine learning is going to be a very big area, with every sector of the economy having the equivalent of a Google self-driving car.

You sound a bit like Ray Kurzweil, saying computers will be the size of a blood cell in 25 years and that robots with computers will live inside our bodies.

Everyone thinks Ray Kurzweil crazy. If he would only stop talking about immortality and ‘this is how I’m programming my body.’ But he has [expletive] called it. Everyone read [Kurzweil’s book] “The Age of Spiritual Machines” ten years ago, and now we have the self-driving car. In my opinion, people are going to look back at what Google has done as the turning point of mankind.

Remember the invention of the transistor? I’m sure at the time there were [trade] articles about it on page 80. Sometimes, these things don’t see so remarkable at the time, but I think machine learning is going to be huge.