Thanks to a pile of new code written in Budapest over the last 3 months, advertisers can now run multiple ad versions in a single day, assigning a weighting to each version. Advertisers can use this feature to keep ads fresh for readers or to test which version is most effective. (If you’re a current advertiser, click “versions” link on an already purchased ad and you can start juggling creative!)

We’ll be letting advertisers know about this and updating our Youtube overview of Blogads, but if you’re a blogger, feel free to post about this new feature to get the word out!

Funny that politicians and economists never do A/B testing of policies.

Sanford Bernstein analysts think Twitter “will likely have to operate it at a loss in perpetuity, or until the next cool Web 2.0 social networking concept comes along and Twitter tweets no more.”

That’s idiotic. Twitter has clearly become a foundation for a new type of communication. The easy, obvious (and perhaps only) route to monetize twitter is to charge users a nominal fee for “registering” their usernames. (Much as domain registrars charge $10-$20 a year for each .com or .org registration.)

Twitter would give the first six months or 300 tweets away free, locking users into the twitter ecosystem.

Twitter now has atomic-powered network effects: who wants to re-follow 1000 people in a new network… or try to convince 500 followers to move elsewhere? Shaq isn’t going to ask his 300k followers to migrate. Perez isn’t going to take his 230k followers elsewhere. I’m not going to want to re-follow 410 people. And all the companies who have built tools around Twitter’s API aren’t going to be eager to recreate those services around other APIs. Once hooked up/in, people will gladly pay $5 a year or $1/month rather than lose their entire network of relationships.

Assuming 50% attrition from its current 7 million users, that would generate roughly $20 million a year… plenty of revenue for servers and the company’s 40-odd staff. It wouldn’t be glamorous, but it would secure profitability and let twitter continue growing the plumbing for a new type of communications revolution.

Update: John Hawkins argues that Twitter’s hordes of followers are inflated by bots.

Update 2: I ran this idea by a friend in-the-know and got what amounted to a “no comment.” Signal or noise?

Keith Olbermann: “Everybody knows I take my instructions from our next guest, the founder and publisher of Dailykos.com, Markos Moulitsas.” (At 2:08)

Bloomberg:

The 8.875 percent 30-year bond that the government sold in February 1989 yields 0.44 percentage point, or 44 basis points, more than the current 10-year 2.75 percent note, which is the most comparable security and yields 2.89 percent. That amounts to about $44,000 a year in interest on a $10 million investment.

In laymen’s language: current buyers of US treasuries are not long-term investors, but flight-to-quality money-stashers. Not a sound base for an rate curve and a sign of much higher rates to come. And higher rates mean fewer home buyers & slower economy.

Heading out for dinner in Austin. Looks like tons of great folks are in town.

Madoff “lost” or “stole” $65 billion. That’s the headline of the day.

How is that possible when Madoff hasn’t even been trading for at least 15 years?

More likely Madoff “redistributed” $65 billion in the form of dividends and cash distibutions to investors.

Or, the bulk of the $65 billion is bogus paper profits. But you can’t really lose money that never really existed, right? (If I wrote you a check for $5 million and it bounced, would we say you’d “lost” $5 million?)

Ads make TV more pleasant: First:

In one experiment, Nelson, along with Tom Meyvis and Jeff Galak of New York University, had 87 undergraduates watch an episode of the sitcom “Taxi.” Half watched it as it was originally broadcast, with commercials for the Jewelry Factory Store and the law office of Michael Brownstein, among other ads. The other half watched the show straight through, without commercials.

After the show was over, the students rated how much they enjoyed it, using an 11-point scale and comparing it with the sitcom “Happy Days,” which they were all familiar with. Those who saw “Taxi” without commercials preferred “Happy Days”, but those who saw the original show, Jewelry Factory Store and all, preferred “Taxi” by a significant margin.

In similar experiments, using other video clips and a variety of interruptions, the results were the same: People rated their experiences as more enjoyable with commercials, no matter their content, or other disruptions. The effect wasn’t limited to watching TV; interrupting a massage also heightened people’s enjoyment, one experiment found.

The opposite was true for irritating experiences, like listening to vacuum cleaner noise: A break only made it seem worse, they found.

Second