How Meta will fail at prediction markets

Meta still doesn’t control a customer-facing platform, and the prospects for pervert glasses don’t look especially good right now, either. They’re losing the same cultural debate that Google Glass did. ICYMI: Wetherspoons bans Meta Glasses in pubs amid privacy concerns.

So what’s next? Prediction markets, apparently. Yes, I admit, predicting that they’re going to fail at this is a safe prediction. Meta trying to build an end-user platform is like Wile E. Coyote trying to catch the Road Runner.

• 2013: Facebook Home
• 2014: The Metaverse
• 2017: Diem (digital currency)
• 2018: Meta Portal

Meta bought Instagram and Whatsapp, and continues to grow as part of the advertising duopoly, but making new consumer-facing stuff, not so much. Open Compute Project was a huge success inside the IT department, but nobody has their hardware at home. The interesting part here isn’t that Meta is going to fail at prediction markets, it’s how they’re going to fail.

“Insider trading”

A lot of carpal tunnels have been sacrificed to warning about “insider trading” in prediction markets. But that’s not really the main problem. There’s no bright line between prediction markets and incentivization markets. At some point, the money at stake gets to be enough to attract a trader who can move the market. In Trading on Violence, Rajiv Sethi writes,

Someone with inside information (or with the capacity to directly influence the outcome) can trade in ways that sharply raise the price prior to event realization. Profits may not be as high as they would be under binary resolution, but they could still be substantial.

So-called “insider trading,” for prediction markets, is a sign of under-informed traders or problematic power centralization. One current big problem is the wildfire markets. A trader who places a bet that no forest fire is going to happen in a certain location is basically paying someone to start a fire. As a money-making strategy, that’s about as good as trading on LIBOR (my friends and I are going to meet up and pick a number, how about you make a bet with me on what the number is) and buyers of “no fire” contracts are either going to lose their money and quit, get arrested for paying someone to start a fire, or both.

Jim Bell’s Assassination Politics envisioned markets as a check on state power—somehow, people would pay for contracts on key government employees—but really, the ROI of replacing one bureaucrat with an ambitious subordinate is much less than the ROI for removing a celebrity tycoon. In practice, the effect of prediction markets will be similar to the effect of the rise of personality-driven publicly traded stocks—more regulation and more corporate governance overhead, to protect high-profile execs by reducing the incentives to remove them.

Oracles as vulnerable points

In some cases, the easy way to move the market is to make an event happen. But another way is to rig the measurement, either physically as when someone set up a hair dryer to blow on a weather station to win a temperature market, or by bribing or threatening the “oracle” or referee. Brad Lipton and Toyosi Odusola write, in How Prediction Markets Are Shaping Real-World Events and Eroding Public Trust,

Although most Americans have never participated in a prediction market, these increasingly popular platforms are already shaping our world more broadly. We are seeing prediction markets entangled in everything from news coverage to war to weather forecasts, with journalists being threatened and battlefield maps literally being redrawn to affect the outcome of prediction market bets.

Prediction markets that free-ride on existing services for oracle are already having trouble. You have to pay the oracle to get a reliable result, and except for a few government-run services (at least for now) the news and research organizations being free-ridden on as oracles aren’t provided with the resources they would need to fill the role. An oracle, for a prediction market, is like a moderator for a social site. It’s the hard part. And Meta dogmatically refuses to staff up and pay for moderators—even when the result is subjecting users to CSAM. I’m not counting the ‘untrustworthy’ social media pushed on some users, or the Meta Caught Paying Nazis to Post on Facebook issue, since it’s likely that Meta decision-makers are just boosting their own politics and the algorithms are working as intended. Victor Tangermann asks “Is Mark Zuckerberg Actually TRYING to Destroy Meta?” and the answer is probably (1) no (2) slop is a political project Anyway, Meta decision-makers would rather fail at anything than lose a labor-management conflict with their moderators. And the same thing is going to happen with oracle failures at their prediction market.

Bonus links

Can public media build its own social network? Chicago Public Media teams up with New Public to try By Sophie Culpepper. For a while now, publishers everywhere have been confronting the bleak reality that, with more and more people turning to social media for their news, news websites — like newspapers before them — are at risk of becoming obsolete. But what if a news website could be something fundamentally different? What if it could beat the platforms at their own game and offer people sick of doomscrolling something better and more useful? (via Chicago’s social network is a huge bet for news. I think it’s the future. by Ben Werdmuller. People trust human relationships, not brands; hosting a space to support those relationships, and forming them authentically and organically with a wider community, will lead to more representative journalism, stronger trust from the public, and better information overall.)

Meta can’t stop states’ $1.4 trillion lawsuit from going to trial by Jon Brodkin. Although Section 230 limits liability for online platforms that host third-party content, Congress didn’t give companies an explicit guarantee that they will not face trials, the 9th Circuit judges’ panel said.

Meta glasses banned from courts in England and Wales by Matthew Weaver. (Places that don’t ban these things will increasingly be seen as unsafe or edgy bros only—and I doubt that edgy bros go out to snap covert pix of each other.)

Judge rules Meta caused “public nuisance” and must fund mental health treatment by Jon Brodkin. (Will teen mental health cases get more parents to ask for help with “parental control” features on smartphones, or make feature requests?)