The Reverse Turing Test

I’m always fascinated by the subtle ways in which our otherwise insightful predictions about the future are nonetheless totally wrong.

The famous Turing Test is paraphrased as “can you tell it’s a robot?” In the ad tech business (and many others) we spend much more time on the problem of “Can you tell it’s not a robot?”

Bitcasa, snake oil, etc

I’ve watched with some amusement the controversy between CrunchFund and TechCrunch and whether Mike Arrington can successfully / fairly cover the startups that he writes about / promotes. On the general question, I have no idea, and I do think Mike has unfairly taken the brunt of the firestorm here when lots of other publications or their parent companies have investments in venture funds. (The NY Times and Giga Om among them.) Also AOL not just approved of the situation, they invested $10m in the fund, so it’s a little disingenuous to give Mike a hard time about it now.

But on the other hand, let’s take a specific case:

Here are some great quotes from the article:

But Bitcasa is not like any of those services. It doesn’t move files around. It doesn’t sync files. It deals in bits and bytes, the 1′s and 0′s of digital data.

I’m sure I shouldn’t make fun of technology reporters who obviously don’t understand anything about technology. Here’s another great quote:

And because the data is encrypted on the client side, Bitcasa doesn’t even know what it’s storing. […] Even if the RIAA and MPAA came knocking on Bitcasa’s doors, subpoenas in hand, all Bitcasa would have is a collection of encrypted bits with no means to decrypt them.

While this is highly unlikely, it is technically possible. However:

Sharing files via Bitcasa is simple too: just copy and paste a file’s or folder’s link (a URL, available on right-click)

One wonders how Bitcasa can deliver an unencrypted version of the file or a URL that contains one if they don’t have the secret key. And finally there is this:

Using patented de-duplication algorithms, compression techniques and encryption, Bitcasa keeps costs down (way, way down, but that’s it’s secret sauce)

Oy. “Patented” (I’m guessing they mean patent pending) proprietary encryption algorithms that somehow let you de-dupe files without knowing what files you actually have. This seems almost certainly like snake oil. You could be comparing encrypted files if all your users share the same encryption keys but of course this means that it would be very easy to figure out which files each user was really storing and the benefit of encryption is lost.

That a startup would be making somewhat overblown claims is not so surprising to me, but what bothers me is the breathy press-release quality of the post. And at the end of the post you see it:

Disclosure: CrunchFund is an investor in Bitcasa. 

It turns out Sarah Perez claims she didn’t know Bitcasa was an investment of CrunchFund when she did the interview. I’m assuming that’s true. But the optics are certainly bad. Imagine a hypothetical meeting where Mike Arrington says, “Hey Sarah, will you interview these Bitcasa guys? They seem great.” Sarah might get positively biased without even realizing that she’s talking about a company her boss has invested in.

Certainly this kind of situation makes me less likely to take what TechCrunch writes seriously.

Apple + OnLive?

Big software companies like Google, Microsoft, Apple and even Facebook have a ton of cash on their balance sheet. Whenever investors start pressuring them to dividend some of that cash to shareholders, the ready answer is always that they’re hanging on to it to facilitate major acquisitions. Steve Jobs said recently justified his $51B cash pile by speaking to the possibility of strategic acquisitions.

Here’s an idea: Apple should buy OnLive. If you’re not familiar with it, OnLive was founded by Steve Pearlman, the guy behind WebTV (acquired by Microsoft) and Moxi (acquired by Paul Allen’s Digeo) and a key figure behind General Magic. (IPO in 1995). OnLive offers it’s customers the chance to play cutting edge games on relatively low end hardware without installation by running the game in the cloud. The user software decodes a streaming video of the game action and sends the user input back up to the cloud. The original announcement was met largely with derision, claiming that round trip times would render the games unplayable.  That talk has since faded with the launch of the service. Clever engineering and placement of servers at the edge of the network have yielded a completely playable product. Users don’t need modern hardware or to install games, they are immediately playable and booted on-demand in the cloud.

Although the game selection is currently somewhat limited the service has several characteristics likely to be attractive to publishers: Complete immunity from piracy, reduced revenue erosion from the used game market and zero distribution costs. Time-based “PlayPasses” and unlimited subscriptions for older games offer a price-point for everyone.

OnLive recently introduced a “MicroConsole” for $99. Since all it needs to do is decode video and stream USB input back up to the cloud, the hardware is cheap. As games improve and require more sophisticated hardware, consumers will not need to upgrade – all the upgrades happen in the cloud and only for the games that require it. A parent faced with a Microsoft or Sony console in the $300 range or a $99 console from OnLive with games of similar quality will have an easy decision.

Question: Which company totally dominates mobile gaming? If you answered Nintendo or Sony, you’re a few years out of date. The answer is Apple. The iPhone and iPod touch enjoy huge penetration, a massive library and an easy payment mechanism. Steve Jobs recently claimed that Apple owns 50% of the mobile gaming market. (A figure disputed by Nintendo.)

However in the “living room game market” (what we once might have called the “console game market”), Apple has no solution at all. In theory, they could do what Microsoft did, burning billions of dollars making custom hardware, but why would they? Apple already introduced an inexpensive living room device that decodes video and retails for $99. It’s called the AppleTV. Customers rent videos through the device and stream them over the net. 10s of millions of customers already purchase games and other digital goods from Apple for their iPhone and iPod Touch. Adding games to AppleTV would only require the addition of USB ports or wireless. Certainly OnLive is going to make the move in the other direction – the console controller conspicuously sports video controls even though OnLive doesn’t currently offer video.

Might Apple try to just build something like OnLive themselves? Sure. But OnLive has built some seriously impressive technology, including tough-to-duplicate leases for space at the network edges, not to mention some fearsome looking IP.

Another interesting tidbit: Steve Pearlman’s employer from 1985-1990? Apple Computer.

Credit: Christian Romming, a co-worker of mine helped develop a number of the ideas above.