Welcome back to Sharp Text. If you’re reading this via email, thank you for subscribing. For anyone reading on the web who would like to receive weekly posts, you can sign up here.
As politics becomes its own genre of pop culture for an entire generation of Americans, laying into the editorial choices of The New York Times has become a national pastime for both the left and right. It was the right’s turn this week, along with many of the most powerful executives in tech, as the Times ran a story on Trump’s AI and Crypto Czar, David Sacks, that was headlined, “Silicon Valley’s Man in the White House Is Benefiting Himself and His Friends.”
Sacks, for his part, responded on X by posting a six-page letter from defamation attorneys who have previously repped the Sackler family, Project Veritas, and former Harvard president Claudine Gay, while Sam Altman wrote, “David Sacks really understands AI and cares about the US leading in innovation. I am grateful we have him.” Indeed, the Valley’s collective response to the article turned into multiple days of “OpenAI is nothing without its people”-style tributes, but for David Sacks and the federal government. On the other end of the spectrum, reporters wondered about the back-channel text messages that spawned the anti-Times outpouring from venture funds the world over, while The Verge observed the groundswell of Sacks support and ran a story headlined, “Silicon Valley is rallying behind a guy who sucks.”
Again, for better or worse — almost certainly worse! — these sorts of episodes are part-entertainment now, generating all kinds of shoulder programming for everyone involved. Sacks’ All-In Podcast will undoubtedly have one of its best weeks ever when the post-Times episode is released this weekend.
With regard to the story itself, past the “Benefiting Himself and His Friends” headline, the body outlines a variety of Sacks-led policy initiatives, but nothing in the reporting ever supports the headline’s implication of nakedly corrupt self-dealing. I think writer Timothy Lee put it best, noting soberly that the story “unhelpfully conflates David Sacks promoting (1) policies that benefit AI companies generally, and (2) policies that benefit specific AI companies Sacks invested in. The examples are almost all (1), but written like they are mostly (2).” I’d add that the story is assiduously not defamatory in a legal sense, but is more pernicious in its perfectly legal smears by conspiratorial implication. What results is a 3,000 word feature that would be misleading to anyone who doesn’t follow the AI and chip policy space (my mom, and most of the Times core audience) and insulting to those who do (people who understand, say, that David Sacks’ chip policies may be controversial, but he does not personally benefit from more Nvidia sales to the Middle East or China).
Here, I’ll stop short of outright sympathy for Sacks. While I’m a lapsed All-In listener, I’ve heard enough episodes to know that he has deployed similar attacks on all kinds of liberal targets, often relying on the same sort of half-baked circumstantial evidence the Times had on offer. And in any event, most of the conversation around this story, like the story itself, ignores the aspect of Sacks’ tenure that I find most interesting. The Times is just asking questions, wondering why David Sacks wants to work in the government, and how he might be benefiting. Sure. But the better question is, why might the government need to employ someone like David Sacks?
Sacks is a “special government employee” who’s limited to working for the government for 130 days per year. He had no prior experience in government. He takes no salary. He is subject to many of the same ethics requirements that apply to full-time cabinet members, but the restrictions are relaxed in certain areas. According to White House ethics waivers, he divested $200 million of his AI and crypto assets before assuming the post that was created for him by Trump. The Times cautions that he hasn’t divested all of those holdings, the timing of the divestments is unclear, and Sacks still retains stakes, through Craft Holdings, his VC firm, in hundreds of hardware and software-focused companies that may benefit from a more business-friendly AI policy set by the White House.
The context the Times omits is that given the investment environment surrounding AI over the past two years, it’s likely Sacks would be wealthier on paper today had he simply never divested any of his holdings and continued to serve in the private sector.
Zooming out, a similar calculus applies to almost any high-achieving private citizen who’s considering government service. It’s a secondary effect of the K-shaped economy that often goes underappreciated. As corporate profits have compounded over the last forty years, the difference between private and public salaries has become progressively more dramatic. A Supreme Court justice makes around $300,000, for example, while a federal judge at the district level is paid $247,000. There are two dozen corporate law firms that pay ten times those numbers to top partners, and many of them pay first-year associates close to what a federal judge makes. The average staffer at the Bureau of Industry and Security, the Commerce Department subgroup overseeing chip controls, makes $107,366. How much more money might those employees get from Nvidia, where the parking lot is full of lime green Lamborghinis? And, of course, juxtaposing the salaries of financial regulators or Treasury employees with those of the executives they regulate has been a laughable exercise for 40 years.
Those dynamics are not new, but they are especially extreme in tech, where working in government not only means making far less money, but also that you operate 3,000 miles away from all the most interesting people in the field, and likely forego future opportunities as a result. Meanwhile, what is new today, and what may not be fully internalized by New York Times readers, is that many of the most urgent questions facing the future of the western world relate to technological shifts that have overwhelming political (and geopolitical) implications. Among them:
- How should the U.S. government incorporate cryptocurrency and stablecoins while preserving both economic stability, regulatory transparency, and the dollar’s global primacy?
- Who should we sell advanced AI chips to, and what precisely are we trying to accomplish with chip controls?
- How might the government effectively regulate tech monopolies to prevent a monopolist in one market using its advantages to distort the outcomes in an entirely new one?
- How should self-driving cars be regulated?
- What guardrails should be put in place to prevent AI disasters and/or mass job loss?
- Are there ways to regulate AI without handicapping innovation and imposing regulatory burdens that favor big companies?
- What sort of policies might delay a war in Taiwan (which would arrest all AI progress for five to 10 years and erase half the stock market’s value overnight)? What’s the most effective way to revive domestic chipmaking capacity?
There are two reasons charting an effective path forward on any of those fronts will require leaning on people outside of Washington. First, to the extent the U.S. wants to capitalize on the opportunities in an industry that has been the envy of the rest of the world for the past 30 years, it helps to have leaders who don’t reflexively resent tech (rare on the East coast). And second, answering any of the questions above demands balancing domestic and geopolitical priorities alongside industry impact and the secondary consequences of any policy, all of which requires granular knowledge of the technologies, leaders and companies at the center of all this change. With AI, specifically, and tech generally, that kind of nuanced expertise eludes most everyone who has spent their career in academia or government. And perversely, that lack of expertise often manifests in tech policies that are hopelessly over-engineered and counter-productive to long term American interests, like last year’s baffling AI Diffusion Rule (which Sacks helped repeal).
None of this is to suggest the government should cede all its decision-making to the Valley, or be absolved of the responsibility to perform comprehensive conflict checks and follow the law with respect to executive branch appointments. Nor am I joining the chorus of tech luminaries saying thank you, David Sacks. I’d like to leave room here for the possibility that the Times story was a mess, VC twitter is awful, and Sacks policies are a mixed bag.
Still, if a venture capitalist’s influence on government policy is depressing to you, I get it, but I would caution that the practical alternative to that AI leadership was Kamala Harris as AI czar, and a 36-page AI executive order under Biden that was grounded in paranoia and equity concern (and partially inspired by Mission: Impossible — Dead Reckoning). There was also the aforementioned AI Diffusion Rule, which introduced three tiers of global chip access, along with nine acronyms (“AIA, ACM, LPP, DC VEU, UVEU, NVEU, TPP, ACA”), while proposing to throttle the flow of U.S. chips and software to more than half the world (creating an enormous gap for Huawei to fill, particularly among U.S. allies in the Middle East).
American tech policy that effectively navigates the years ahead will require reckoning with the reality of modern government’s limits, as well as the multi-dimensional complexity of its priorities. Rather than casting Kleptocratic aspersions on a part-time government employee with so much industry experience that the appearance of conflicted interests is effectively unavoidable, a more useful Times story might have identified those dynamics and asked whether arrangements like Sacks-as-AI-czar may nevertheless become more common in administrations of the future. Provided there are strong national security and regulatory voices that remain in the room, it seems clear the U.S. government incorporating genuine tech expertise is better than the alternative.
What I’m Reading This Week
- Part 1: My Life is a Lie. A great piece on the limits of available economic statistics, the need for a new definition of poverty, and the “valley of death” among working class Americans that breeds resentment and social instability. Some of the numbers in this piece have been questioned as it’s gone viral, but the structural dynamics read as deadly accurate. (Part 2 was also excellent).
- My Boy. A fantastic and moving piece on raising an autistic son, from my friend Ethan Strauss.
- How NBA Legend Michael Jordan Is Blowing Up NASCAR’s Monopoly. “Several jurors were dismissed because of their love for Jordan, whereas another was kicked off the case because of her dislike for one of the driving teams involved. And then there was the guy who joked on his juror form that his hobby is ‘heavy drinking;’ he was ultimately chosen to serve.” Wonderful stuff.
- MAGA’s Aquatic Valkyrie. A comprehensive response to an episode of Pablo Torre’s podcast that seemed pretty naive about how political activism works.
- Stanford Earth Sciences Chair Collaborates with China’s Nuclear Program. An investigation of a Stanford scientist’s apparent collaboration with the CCP nuclear program; this is one of dozens of examples with respect to the academy’s endemic China problems. Part 1 in this remarkable series from student journalists was similarly disturbing.
- China is making trade impossible. For anyone new to the world’s trade tensions, this 900-word op-ed is the most cogent statement of the problem (and why, particularly for Europe, there are no good solutions).
And with that, we’re done! Thanks for reading, and have a great weekend.
Sharp Text is extension of the Stratechery Plus podcasts Sharp Tech, Greatest of All Talk, and Sharp China. We’ll publish once a week, on Fridays. To subscribe and receive weekly posts via email, click here.