At this point, it’s pretty clear what Donald Trump wants from Mark Zuckerberg. But what does Zuckerberg, who has now gone to Mar-a-Lago twice since the November election, want from the President-elect?

That’s the question I’ve been asking sources in and around Meta over the last several days. They all described Meta’s relationship with the outgoing Biden administration as incredibly hostile. It’s safe to assume that Zuckerberg wants a reset for the MAGA regime, especially since Trump threatened not that long ago to imprison him for life. 

In Trump’s America, removing tampons from the mens’ restrooms on Meta’s campuses, — a real thing that just happened — is as much a business decision as a political one. Destroying ‘woke’ ideology is a key pillar of Trump’s stated mandate. Others who know they need to play the game, like Amazon, are also starting to fall in line. Even still, Zuckerberg is transforming Meta for this new political reality at a speed that’s unusual for a company of its size and influence. Founder mode.

In his conversation with Joe Rogan and his video on Instagram, Zuckerberg shares a laundry list of issues that Trump could help him with: fighting other countries that are ratcheting up their policing of his platforms, stopping Apple from dictating how he builds mobile apps and smart glasses (the latter is increasingly important to Meta’s future), and, perhaps most importantly, keeping domestic AI regulation from slowing his efforts to crush OpenAI. Elon Musk has bought Trump’s ear. But the more time Zuckerberg spends in Mar-a-Lago, the more Sam Altman and Tim Cook should be worried.

Then there’s the US government’s case to break up Meta that’s set to go to trial in a few months. After the blur that was the last four years, it’s easy to forget that this lawsuit was filed at the end of Trump’s first term by a Republican FTC chair, not Lina Khan…

Most of the headline reactions from the past week have focused on Zuckerberg’s decision to end Meta’s third-party fact check program. It was a convenient scapegoat for company executives that, frankly, never lived up to the goal of bringing more neutrality to Facebook and Instagram. The Community Notes alternative Meta is cribbing from X was not on the product roadmap before this week, so it will probably be awhile before everyone sees it in the wild. 

The announcement that US moderators would be moved from California to Texas is perhaps the most cynical of them all; talk to anyone who knows and they’ll tell you the vast majority of moderators are already based in Austin.

The hateful speech that is now allowed on Meta’s is eye-popping on its face and will be deserving of more scrutiny in the coming weeks. The decision to start recommending political content again is a 180-degree turn for Zuckerberg. But insiders believe that the most impactful change for users of Meta’s apps will be the softening of its systems that remove content for potential policy violations. 

Out of all the announcements Meta made last week, this is the one I believe is the least connected to Trump. Meta execs have been signaling for a while that they know they are mistakenly removing too much content that doesn’t actually break the rules; I’m told it’s one of, if not the, biggest complaint in user surveys. If done correctly, dialing back on moderation mistakes may be the only thing Zuckerberg announced that makes everyone happy.

• CES is for dealmaking now: Each year, the official CES show — the sprawling show floor and flashy keynotes — feels more like an advertising exercise and no longer a place to launch real products. Most of the energy has moved to private meeting rooms and happy hours at the Wynn, Aria, and Cosmopolitan, where tech execs are schmoozing CMOs and getting deals done with partners all week. At this shadow CES, everyone seems to agree that the show is more alive than ever. Booths on the show floor have become marketing tools to show clients before you take them to a steak dinner. The challenge for the organizers of CES will be figuring out how to bridge the growing influence of this part of the show with their current business model of charging people to walk around booths filled with smart toasters and concept cars.
• TikTok may just get banned: Imagine an alternate world in which the Chinese government is about to ban Instagram from operating in the country and Mark Zuckerberg is in hiding. That’s the situation with ByteDance and its founder Zhang Yiming, who stepped down from the CEO role after the last US ban attempt but still controls the company. He let TikTok be banned in India and seemingly has no interest in the app surviving this time, so why wouldn’t he let the same thing happen again?
• Google and OpenAI flick at what’s next: Google’s DeepMind unit is starting “an ambitious project to build generative models that simulate the physical world,” which it believes “is on the critical path to artificial general intelligence.” Meanwhile, OpenAI is returning to its early roots by starting a “general-purpose robotics” team that will build hardware and push “towards AGI-level intelligence in dynamic, real-world settings.” We may have hit a scaling wall on text data but the big labs clearly see an opportunity in 3D. (See also what Nvidia announced last week.)
• Other headlines you may have missed: Tencent (a large investor in Epic Games, Snap, and US tech companies) was put on the Pentagon’s blacklist for being allegedly under the influence of the Chinese military. Tim Cook’s total compensation rose 18 percent last year to $74.6 million. Elon Musk is hosting an inauguration party for Trump in DC with Uber and The Free Press. Sam Altman’s sister filed a sexual abuse lawsuit against him.

Some recent, noteworthy job changes in the tech world:

• A bunch of changes at Meta: UFC CEO Dana White, Exor CEO John Elkann, and Charlie Songhurst joined the board. Joel Kaplan is running policy and comms now. After a stint at Google, I’m told Michael Levinson is coming back as VP of product for the Integrity org. (Good luck!) Head of civil rights, Roy Austin, is leaving. And former DEI chief Maxine Williams is now head of “accessibility and engagement.”
• Elon Musk’s X named a couple of new leaders: Romina Khananisho is the new head of government affairs and John Nitti is head of “ad innovation.”
• Calista Redmon joined Nvidia as VP of “global AI initiatives,” where she’ll “drive adoption of the NVIDIA platform for national and regional AI initiatives.”
• Sophia Dominguez, Snap’s director of AR platform, is leaving.

If you haven’t already, don’t forget to subscribe to The Verge, which includes unlimited access to Command Line, all of our reporting, and an improved ad experience on the web.

As always, I want to hear from you, especially if you work at Meta. Respond here, and I’ll get back to you, or ping me securely on Signal.

Las Vegas is punishingly dry. The arid winter air means I woke up on Day 3 of CES 2025 with a nosebleed, chapped lips, and ashy legs. This in spite of the fact I slathered myself with two pumps of a fermented bean essence, eye cream, moisturizer, and a lip mask. Staring at my face in the hotel mirror, I wonder if any of those products were doing what they’re supposed to — and if maybe, I should try something different.

This is why I was so eager to try L’Oréal’s Cell BioPrint.

For anyone who’s struggled with their complexion, the Cell BioPrint feels like a holy grail gadget. The device is a mini-lab setup that analyzes a skin sample to generate a report about your skin’s current condition. It’ll also “grade” your skin with regard to oiliness, wrinkles, skin barrier function, pore size, and uneven skin tone. Based on the proteins in your skin, you’ll also see whether you’re more likely to be susceptible to those issues down the line — even if they aren’t issues now. The test also determines whether you’re responsive to retinol, a popular and well-studied skincare ingredient that nevertheless causes a ton of confusion online.

In my demo, using the Cell BioPrint was simple. Mainly because I didn’t have to do a thing. While L’Oréal hopes to one day make Cell BioPrint an at-home device, it’ll first be aimed at retail, dermatology offices, and skincare clinics. As in, the actual test will be done by a professional to ensure accuracy. A L’Oréal staffer collected a sample from both my cheeks using a special kind of sticker. I watched as he then dissolved it in a buffer liquid, placed the resulting solution into a cartridge, and fed that cartridge into a machine. I also had my cheeks and forehead scanned with an imaging wand before answering two questions about my age and demographic data.

After a few minutes, I got to see my results. Apparently, I’m doing something right because the report said that my chronological and biological age were aligned. But it also said that while my skin barrier function is currently good, I’m biologically prone to have issues with that as I get older. My report also said that I don’t have to be concerned with pore size — either now or in the future, and that I am highly responsive to retinol, meaning my skin is likely to tolerate it well.

There were several other insights I won’t bore you with, but after going over my results, I have a better sense of what I need to focus on. For instance, I should continue using moisturizers with ceramides, button up my sun care habits outside of sunscreen, add a vitamin C to help with my skin tone, and add a retinol. I also know what I don’t need or can remove from my routine. For example, I don’t need to buy products targeted toward shrinking pores.

Most of this wasn’t surprising. My results aligned with the concerns I’ve noticed and where I tend to concentrate my skincare routine. For example, the few times I’ve tried retinol, I’ve never had an hint of the sensitivities other people seem to experience.

Of course, this kind of personalized recommendation is only as good as the science behind it. After all, lots of health and beauty tech companies promise the moon, but are intentionally coy when it comes to explaining why you should trust them. Barring regulatory certification (which isn’t generally required for wellness and beauty tech), it’s left up to the individual to try to discern whether an explanation passes the smell test.

With that in mind, I asked L’Oréal to dive deep into the science — which Guive Balooch, global vice president of L’Oréal’s Technology Incubator, happily obliged.

Balooch says the company had 800 biologists working (and publishing clinical studies) to discover whether there are specific proteins expressed in skin cells that could determine your risk factors for certain skin conditions. The challenge, he says, is that there are a lot of proteins produced by the body. Finding the few out of the thousands that can give actionable skincare insights is like finding a handful of needles in a giant haystack. It required L’Oréal’s researchers to sequence all of them and then find the relevant biomarkers.

This particular branch of research, Balooch says, is called proteomics — or the study of how proteins are expressed in the body. “It’s understanding that our cells every day are making proteins. Depending on our lifestyle, our geography, and our genes, they will make more or less of these proteins. It changes over time and habits can change it as well.”

Balooch says L’Oréal tested 4,000 people over 10 years across the US, Europe, South America and Asia, and found five proteins related to skin health, including two that are related to how well a person’s skin will respond to retinol. And while the Cell BioPrint can only currently analyze responsiveness to retinol, other ingredients like niacinamide and hyaluronic acid are also in the works.

Cynically speaking, a device like this could also be viewed as a pseudoscientific way to sell more products to a group of people already prone to reach for their wallets. But Balooch argues the Cell BioPrint isn’t meant to encourage people to buy more.

“In some ways, it’s meant to tell people what not to buy,” he says. “Of course, we would love to sell more products, but not by overconsumption. That’s not actually good for your skin. It’s about helping people find the right products based on science.”

To that end, Balooch has a point. The skincare market right now is rife with misinformation and influencers pushing expensive 10-step routines that sometimes do more harm than good. I know better, and yet I’m as guilty as the next skincare nerd. (I’m still haunted by the full bottle of snail mucin influencers claimed would fix all my problems. It broke me out instead.) Even now, I know that I’ll probably get influenced again. The difference is at least I can choose to be influenced towards a retinol cream or vitamin C serum — things that have a higher chance of helping me — rather than throwing my money to the wind.