It’s not just A.I. companies begging for more regulation

One of the most interesting speakers here is Miki Kuusi, cofounder of Finland-based delivery firm Wolt, which DoorDash bought for $7.5 billion last year. Kuusi is now both Wolt’s CEO and DoorDash’s head of international and would like to see more regulation of the gig economy. Why? Because it allows Wolt to treat its people well without being effectively penalized.

“We want to do the right thing, so we’re self-regulating,” Kuusi said onstage. “What pisses me off like hell is seeing a competitor that’s literally getting a competitive advantage because they don’t do the same thing.”

I caught up with Kuusi afterward to get more clarity on that. “If you look at the courier side of the work, it’s basically a balance between freedom and flexibility—this is why most people want to do this kind of app-enabled work,” he told me. “But you also need to have meaningful earnings for the work you do, and you also need to have protections and safety nets if something goes wrong.

“We regulate ourselves when it comes to making sure the earnings are high enough. We see competitors that don’t necessarily do that. We self-regulate when it comes to having global insurances for our couriers, which not every company is doing,” he said. “These are things that are in terms of values that are important to us, but I think these shouldn’t be up to values. This should be something that is not just possible to do, but also demanded of every company in the industry.”

That’s a very different tune to what we used to hear from the gig-work sector, which was not so long ago desperate to avoid regulation—not least because being under-regulated allowed them to outcompete traditional rivals that were burdened with loads of rules, such as the taxi industry. 

But it’s a refrain we hear again and again these days, from a variety of tech players. Currently, the most prominent example is the A.I. industry, in which we see leading lights such as OpenAI CEO Sam Altman touring the world, begging for regulation not just at the national but global level.

Given the white-hot level of competition in A.I., it’s easy to take a cynical view of these calls, not least because the giants emitting them have resources to deal with regulatory compliance that smaller upstarts lack. 

That’s a very real problem, but I don’t think it’s the whole story. The likes of Altman and Alphabet CEO Sundar Pichai know they have to contend with a public that deeply, deeply distrusts A.I. That’s going to stifle uptake, and regulation could mitigate that problem. There’s also the age-old factor of certainty: better the regulation you know than the regulation that might give you a nasty surprise down the line.

But I also think A.I. leaders are genuinely concerned about the implications of A.I.’s rapid evolution, and see regulation as a way of avoiding the worst outcomes. Like Kuusi, maybe they want to do the right thing, and need regulation to ensure that rivals can’t force them to take a different route by operating with less noble motives.

The truth probably lies somewhere in the middle of these various imperatives. But one thing is for sure: Regulation is no longer anathema to the tech sector. That’s a sign of a maturing industry with unprecedented significance to society. As Kuusi told me about the phenomenon: “I think it’s about size and scale first and foremost.”

Want to send thoughts or suggestions to Data Sheet? Drop a line here.

David Meyer

Data Sheet’s daily news section was written and curated by Andrea Guzman.

NEWSWORTHY

U.S. users face the end of password sharing on Netflix. Netflix began cracking down on password sharing earlier this year in Canada, New Zealand, Portugal, and Spain. Now, it’s bringing the password police to the U.S. In an update on sharing yesterday, Netflix reassured users that every member of a household would still be able to use its video streaming service wherever they happen to be located. But for people not actually part of the “household” (which Netflix is defining based on IP addresses, device IDs, and account activity) there are two options: Convert your user profile into a separate Netflix account, and pay the standard monthly fee, or pay $7.99 a month to add the non-household-member’s profile to the household account. 

Meta’s latest round of layoffs. Meta is following through on plans announced in March to cut 10,000 jobs over the course of two rounds of layoffs in April and May. This latest round will affect about 6,000 people mostly in business roles following the last round that hit tech teams. The Washington Post reports that some current and former Meta employees are worried that the cuts could hurt its responsiveness to political misinformation, foreign influence campaigns, and regulatory challenges just as dozens of elections are slated for next year. But Meta said it’s prepared and will continue investing in teams and technologies that protect its users.  

Virgin Orbit’s bankruptcy sales. Bankrupt rocket company Virgin Orbit has sold its assets in a string of deals totaling more than $35 million. The deals, which must be approved in federal bankruptcy court, include Stratolaunch’s purchase of Virgin Orbit’s Boeing 747 airplane, dubbed Cosmic Girl, and Rocket Lab USA’s $16.1 million purchase of Virgin Orbit’s rocket factory in California. This follows the company’s decision to cease operations after being low on cash in mid-March and furloughing almost its entire staff while it sought a financial lifeline.

SIGNIFICANT FIGURES

10%

—The portion of equity in Stability AI that a Virginia doctor, who claims to be a cofounder of the company, says belongs to him. Computational biologist Tayab Waseem filed a lawsuit against Stability AI and its cofounder and CEO, Emad Mostaque, last week, claiming that from July 2020 through July 2021, he was the chief scientific officer of the company and was promised a stake instead of a paycheck. The complaint was dismissed voluntarily by Waseem’s counsel on the same day it was filed, Vice reports, without specifying the reason.

IN CASE YOU MISSED IT

News website Boring Report is using A.I. to change how people consume news—literally making it ‘boring’ by avoiding sensationalism, by Prarthana Prakash

Pentagon attack hoax that triggered brief market selloff illustrates pitfalls of A.I.-driven fake news, warns Deutsche Bank, by Christiaan Hetzner

OpenAI’s Sam Altman and Google’s Sundar Pichai are now begging governments to regulate the A.I. forces they’ve unleashed, by Prarthana Prakash

After 5 shark attacks on Long Island in just a year, New York’s governor is using drones to track any Jaws wannabes lurking nearby, by Chris Morris

Ford CEO Jim Farley promises a new long-range electric SUV so quick he calls it a ‘personal bullet train’, by Christiaan Hetzner

BEFORE YOU GO

Videos created from brain readings. In a paper released last week, researchers said they used fMRI data and the text-to-image A.I. model Stable Diffusion to create a model that generates video from brain readings. Researchers from the National University of Singapore and the Chinese University of Hong Kong say their model is called MinD-Video and is capable of making videos that are “high quality” with respect to motion and scene dynamics.

Vice reports that training the system involved using a publicly available dataset with videos and fMRI brain readings from test subjects who watched them. Researchers published the original video of horses in a field as well as a reconstructed video of the scene. In another video, a car is in motion, while the reconstructed video shows the perspective of someone traveling on a road. In the paper, the authors write that the field “has promising applications as large models develop, from neuroscience to brain-computer interfaces.”