First, people fretted about algorithmic bias. Now, they’re worried about A.I. bias extending that danger. But soon, the big concern may be bias in neurotechnology—a rapidly swelling field that includes the likes of Elon Musk’s Neuralink, and the Jeff Bezos– and Bill Gates–backed rival, brain implant maker Synchron.
The U.K.’s data protection regulator, the Information Commissioner’s Office (ICO), today said it expected to see devices that tap into signals coming from the nervous system to “become widespread over the next decade.” While researchers are already using brain implants to combat dementia and help paralyzed people walk again, the watchdog said, “neurotechnologies are rapidly developing for use in the personal well-being, sports, and marketing sectors and even for monitoring people in the workplace.”
The ICO fears dangerous bias could creep into the models being developed for reading and interpreting neurodata, if the devices aren’t assessed on neurodivergent people in particular. For example, neurotechnology systems could have the baked-in assumption that certain patterns are “undesirable…[so] those with those patterns may then be overlooked for promotions or employment opportunities.”
“Neurotechnology collects intimate personal information that people are often not aware of, including emotions and complex behavior. The consequences could be dire if these technologies are developed or deployed inappropriately,” said the agency’s regulatory risk chief, Stephen Almond.
Synchron is already conducting human trials of its brain implant—it said earlier this year that the implant “appeared to be safe and enabled patients to do hands-free texting, emailing, online banking and shopping, and communicating care needs using their thoughts.” Neuralink, which claims its brain interface will “restore autonomy to those with unmet medical needs today and unlock human potential tomorrow,” finally got FDA approval for human testing a couple of weeks ago.
Meta also dabbled in the brain interface space for several years under its former Facebook guise, but pulled out in 2021, on the basis that consumer head-mounted hardware was a long way off. It decided to instead focus on developing a wrist-worn device that could sense electrical motor nerve signals—the sort of thing that could prove useful for interacting with the metaverse.
The ICO reckons such wearable devices will be the ones that gain real market traction in several years’ time, being integrated into fitness trackers, deployed in the workplace, and used to play VR/AR games.
However, if you’ve seen that fascinating tweet by neurotechnologist Sterling Crispin, who worked on Apple’s Vision Pro headset and described it as “a crude brain computer interface via the eyes”…no, the ICO is not concerned about that. “We are talking about the collection of data directly from people’s brain and nervous system,” a spokesperson told me.
The ICO said it will issue specific neurodata guidance by 2025. In the meantime, it has a report out on the subject, which you can read here.
More news below.
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
Google and Salesforce team up. After an analytics and big-data collaboration in 2017, tech giants Salesforce and Google have decided to partner up again. This latest project will allow companies to use their data along with custom machine learning models to anticipate their customers’ needs. It will involve integrating Salesforce’s Data Cloud with Google’s BigQuery and platform Vertex AI so that users can bring any A.I. models they’ve trained in Vertex across Salesforce’s platform. The agreement follows Salesforce’s announcements this year on new A.I. tools like its assistant for customer relationship management known as Einstein GPT, and as Google trails behind Amazon and Microsoft in the U.S. cloud infrastructure market, TechCrunch reports.
Twitch walks back on ad changes. Twitch’s changes to its branded and sponsored content policy slated to go into effect July 1 banned “burned in” video, audio, or display ads. But after streamers like MrBeast shared frustration over the hit it would bring to their revenue and considered boycotting the platform, Twitch decided to scrap the changes. In a tweet, the Amazon-owned platform addressed streamers, saying it would not prevent their ability to have direct relationships with sponsors. “You will continue to own and control your sponsorship business,” Twitch wrote.
DeSantis ad uses a fake A.I. image of Trump. Earlier this week, the “DeSantis War Room” Twitter account shared a video that includes three A.I.-generated pictures of former President Trump embracing former White House chief medical advisor Anthony Fauci. Digital media forensics experts told AFP that the images were very likely fake and couldn’t be found in reverse image searches. Some experts have worried about the impact advanced A.I. could have on the 2024 election, while others note that human-created misinformation has spread in recent elections so it may not make much of a difference.
SIGNIFICANT FIGURES
6
—Ranking of the EPA’s air quality app on Apple’s app store, ahead of Facebook, Twitter, and Snapchat. As New York City’s air quality suffers from smoke that traveled from Canadian wildfires, the metro has topped the list for the worst air quality in the world.
IN CASE YOU MISSED IT
Even OpenAI CEO Sam Altman thinks people are going a little too crazy over A.I.: ‘It’s wildly overhyped in the short term’, by Prarthana Prakash
‘Father of the iPod’ says Apple has ‘truly jumped the shark’ with its $3,500 virtual reality headset, by Chloe Taylor
Two of the richest crypto bros saw nearly $2 billion wiped off their combined wealth following Binance and Coinbase crackdown, by Christiaan Hetzner
Google ramps up return-to-office push by using in-person attendance as part of employee performance reviews, by Prarthana Prakash
‘Black Mirror’ creator Charlie Brooker says he tried to get ChatGPT to write an episode of his hit show—but it was ‘sh*t,’ by Chloe Taylor
BEFORE YOU GO
Engineers try to beam solar energy from space to Earth. In January, a team at Caltech launched an orbiting suite of tests on components for space-based solar power, and it’s now sending back encouraging results. Others are also making an attempt. U.S. military labs think they might be able to transmit space-based power to bases or combat zones someday. Meanwhile, the European Space Agency has a blueprint for a space-solar network and the China Academy of Space Technology is aiming for a power-beaming satellite prototype by 2028.
The Wall Street Journal reports that a big challenge is identifying a safe and reliable way to transmit gigawatts of power to the ground and then convert it into electricity that people can use. One method could be microwave beams, but the European Commission found the beams wouldn’t be strong enough. Still, it’s one of the most favored methods at the moment. “It’s basically the same technology as wireless charging for your cell phone,” Chris Rodenbeck, head of the Advanced Concepts Group at the U.S. Naval Research Laboratory in Washington, D.C., told the paper.
This is the web version of Data Sheet, a daily newsletter on the business of tech. Sign up to get it delivered free to your inbox.
