The Business of the Future

by Ben McLeay

 

Larry Bekhterev apologises profusely when he arrives at the unassuming cafe he has picked out for us in the heart of San Francisco’s Mission District. He is three minutes late.

Looking stylishly unkempt in plaid and well-worn denim, Bekhterev is an instantly likeable explosion of enthusiasm. Charismatic, effusive, and very, very sharp, it is not hard to understand how Bekhterev—in the space of only two years—went from sleeping in his car through college to being the third richest person under 30 in the world.

“I wouldn’t say I’m a particularly gifted programmer,” he tells me, practically already vibrating with energy as he takes a sip of his coffee. “You couldn’t swing a cat at Google or Amazon or Facebook without hitting someone who can cut better code than me.” He pauses to laugh. “I don’t recommend you swing a cat in there, though. Wouldn’t go over well.”

Whether or not he’s downplaying his abilities is irrelevant, he believes. “I’m where I am now because I saw a problem and I had, in my own small way, the capacity to address it.” That problem, as you likely already know, is one of the most fundamental problems humanity as a whole has wrestled with. And the solution was Bekhterev’s app, Vzn.

#

In 1967, legislation passed by the British Parliament established the Royal Public Registry of Parapsychological Premonition (now simply the Central Premonitions Registry). In 1968, the United States followed suit with the Federal Register of Premonitions, accompanied by a dozen or so state-level bodies. Australia, Canada, Japan, then-British Hong Kong, New Zealand, Sweden, France, Russia, and Spain all established their equivalents within the following decade.

While the mechanism for premonition is still understood in only the vaguest of terms today, in the 1960s it was a complete mystery. The British government knew one thing and one thing alone: with a large enough sample, premonition reliably demonstrated accuracy greater than you would expect from random chance. Once the government’s trained parapsychologists had filtered, sorted, categorised, and interpreted premonitions submitted by the public, a list of likely future events could be collated and distributed to the public, allowing prognostication with an accuracy somewhere between a weather forecast and anticipating a coin flip.

In Britain, Australia, and New Zealand, these took the form of a telephone book–sized directory of premonitions, delivered free to households quarterly, with the first half indexed by date and the second half by location. In the US, citizens could access the registry for a small fee via a human-operated telephone service, which was later automated in the mid-70s.

Sir Robert Jennings, now Chief Officer of the Central Premonitions Registry, was only 19 when he started as a filing clerk in the then-Royal Public Registry of Parapsychological Premonition. “I was there on the day that Prince Philip cut the ribbon on the building,” Jennings tells me in his office, which—with its curios, arcane charts, and precariously balanced stacks of leather-bound books—seems more suited to a 14th-century alchemist than a public servant. “Ostensibly, I was just there to move the properly written-up transcriptions from the clerks’ office to the filing room, but in those early days, we had no systems in place. Despite my youth and lack of experience, I was instrumental in setting up the processes that are still used to this day.”

Some half a century later, save for the digitisation of the process, the CPR functions much as it did when it opened. Premonitions are submitted by the public; properly filed and tagged by clerks; and then authenticated and weighted by trained parapsychological officers. An independent audit of the CPR undertaken in 2015 found that, on average, it took 6.7 days between a premonition being submitted to the call centre and that premonition being listed on the website.

#

“Watch this,” Bekhterev tells me. We’re holding our respective smartphones next to each other as he types a premonition into Vzn. He hits a plus icon at the top right of the app, economically filling in the sparse handful of fields with the speed of someone more comfortable typing on a phone than on a keyboard. Type: Dream. Locus: AT&T Stadium. Temporality: Imminent. Outcome: Giants win.

My phone shows a Google Maps-style view of the area around AT&T Stadium. As soon as he hits ‘submit’, a green circle around the stadium grows slightly larger, simultaneously turning a slightly more vibrant shade of green. “And that’s it,” he says, an unmistakable pride visible on his face.

“We had doubts at first, obviously.” It’s late afternoon, Bekhterev is sitting in an office decorated with spraypainted skate decks and framed posters of 90s west coast rappers, toying with the rolled-up sleeves of his shirt, unrolling and rolling them like he can’t quite figure out the appropriate length. “The app needs numbers to function. It’s a feedback loop: the more people there are, the better it works, but we need it to work well to get more people. We needed that initial influx.

“It worked, though. Maybe it was luck, maybe people were just sick of having to wait a week to hear about things that were gonna happen tomorrow.”

According to Bekhterev, their own analysis shows that Vzn ends up displaying 90% of the predictions released by the Federal Register of Premonitions, and will surface ‘a few hundred’ every week that aren’t documented at all by the FRP. (Independent analysis found that that figure is an exaggeration, but only a very slight one.)

“There are no experts. No one’s at the wheel. No one at Vzn has a parapsychological background. Users post their premonitions, others users rate the accuracy of those premonitions after the fact. Our algorithm turns those ratings into a score that quantifies a user’s clairvoyance or luck or ability to extrapolate current events into future ones or whatever,” Bekhterev says. “That might sound flippant, but it’s irrelevant to us.

“What’s being scored is their reliability based on past performance, it has nothing to do with how they’re doing it. The algorithm doesn’t care, so neither do we.”

***

Read the rest of this story in Creeper Magazine - Issue One coming soon from Oh Nothing Press.

Ben McLeay tweets as @thomas_violence and is part of Australia’s favourite podcast - Boonta vista SOCIALIST CLUB.