This is the web version of Data Sheet, Fortune’s daily newsletter on the top tech news. To get it delivered daily to your in-box, sign up here.
Donald Trump has discovered artificial intelligence. And he didn’t even have to go to Davos to accomplish this feat.
The president issued a directive Monday, formally the Executive Order on Maintaining American Leadership in Artificial Intelligence, that lays out a framework for something having to do with A.I. I say something because after having read the document I’m not altogether clear on what it’s about. (In my diligent primary research I took inspiration from the crafty Alan Murray, who recently read Alexandra Ocasio-Cortez’s “Green New Deal” resolution; I recommend his analysis.)
Many have noted there are no dollar figures associated with Trump’s treatise. Less noted is that there’s not much substance to it either.
Here’s a quick synopsis. The idea is that the U.S. already is great at A.I. and needs to stay that way, especially with the Chinese coming on fast. Trump suggests five A.I. “principles”: 1) A.I. breakthroughs are good; 2) A.I. should have standards; 3) A.I. training is important; 4) A.I. needs to be a trusted technology; 5) foreign markets should remain open to U.S. A.I. He also offers six “objectives”: 1) government should invest in A.I.; 2) federal data should be available for A.I. to crunch; 3) barriers to achieving A.I. should be reduced; 4) someone should develop A.I. standards; 5) workers should be trained to do A.I., including federal workers; 6) the U.S. should have an A.I. “action plan.”
There’s more, including the admonition that government agencies that do R&D should make AI a priority and that agencies that fund education should fund it in A.I. It’s all very swampy, including a reference to something called the “President’s Management Agenda and the Cross-Agency Priority Goal: Leveraging Data as a Strategic Asset.”
It’s fair to assume a management consultant wrote this stuff. It’s also good validation of what it means to have a businessman in the White House: this president knows how to spread around and re-direct stuff that’s already in progress or called something else, slap a buzzword on it, and call it progress. All that’s left to produce is a “magic quadrant” picture to illustrate how great this all is.
It’s not even clear to me the U.S. needs an A.I. policy. Standards and education and data sharing are things government agencies can do without an executive order—and that industry can do too. Presidents who are serious about a national objective do things like create federal agencies, like what became DARPA, an Eisenhower-era confection that led to the invention of the Internet.
Perhaps Trump’s “action plan” will lead to something similar.
YouTube copyright hell part 57: In a new twist on the video platform’s long-running intellectual property troubles, extortionists are filing fake copyright complaints in order to obtain “two strikes” against smaller channels—and demanding a pay-off not to file a third one. Observers note this stems from YouTube reserving its scrutiny for accused infringers rather than those who accuse them.
Working at Apple’s “Black Site”: Tales of contractor misery typically focus on Apple’s overseas assembly plants but you can also find them six miles from the the Cupertino mothership. In a new dispatch on Silicon Valley’s caste system, Bloomberg describes a Sunnyvale plant where Apple contractors carry out software testing in a facility with limited bathrooms and no perks. The kicker—after being enticed in some cases to “work for Apple” the contractor forbid them from listing the company on their LinkedIn page.
More than a trace of DNA: What privacy worries? Despite well-publicized privacy concerns tied to commercial ancestry tests, more people than ever are submitting saliva samples to the likes of Ancestry and 23andMe. In 2018, the number of participants exceeded all previous years and, if the pace continues, companies will have more than 100 million samples by 2021.
Give it me straight, doctor: The problem of misdiagnosis is endemic in medicine because physicians are, well, human. Now, new research culled from a pediatric hospital in China suggests artificial intelligence does a better job at diagnostics. If deployed at scale, A.I. could not only improve diagnosis but help patients where doctors are scarce. And as with most things A.I., developing and deploying such systems is easier in China where researchers can simply help themselves to the relevant data without quibbling about things like patient privacy.
Yesterday’s item about an interview on Google’s Waymo should have been attributed to CTO Dmitri Dolgov. The company has also recently launched a product in the form of self-driving service, Waymo One. We regret the error.
ON THE MOVE
The popular news discovery tool Nuzzel will lose CEO Jonathan Abrams and COO Kent Lindstrom, who are leaving the company after its acquisition by subscription service Scroll… Speaker maker Sonos is losing its CFO, Michael Giannetto, who is leaving at the end of the year.. .Flex, the global tech design and manufacturing firm, has hired energy veteran Revathi Advaithi as its new CEO… Former AOL/Oath CEO Tim Armstrong announced a new company, dtx, that will invest in digital retail ventures.
FOOD FOR THOUGHT
Why don’t media outlets put more emphasis on quality over quantity? What if, instead of cranking out dozens or even hundreds of ephemeral items, news sites only published what matters? It’s a nice idea but it’s also a ticket to obscurity. In a smart how-the-sausage-is-made account, veteran tech journalist Owen Williams explains how his new site has been repeatedly snubbed by Google News—in part because he doesn’t publish often enough. The piece is yet another indictment of the perverse incentives that drive much of today’s media making.
IN CASE YOU MISSED IT
Crypto Detective Firm Chainalysis Raises $30 Million from Accel By Jeff John Roberts
Amazon to Acquire Alexa-Compatible Wi-Fi Mesh Router Company Eero By Brittany Shoot
Verizon Has the Best Mobile Network of Any Carrier By Aaron Pressman
The Fight Over High-Tech Supremacy Isn’t New. We Just Haven’t Learned Our Lesson By Mario Daniels and John Krige
BEFORE YOU GO
The trauma over school shootings has led well-meaning people to make active-shooter drills part of the American educational experience, leading 4.1 million students to experience at least one lockdown or lockdown drill last year, including some 220,000 students in kindergarten or preschool. But to what end? A thoughtful Atlantic piece makes the case we’re inflicting terrifying trauma on our children to prevent what the figures—fewer than 150 children and adults shot to death in U.S. schools in nearly two decades—suggest is an incredibly unlikely event.