Sam Silverman

Today I'm launching Silverman Strategy Group. Technology, policy, and comms are colliding faster than ever. That’s what this firm is built for.

These are the correct questions. To put it nicely, it feels hard to see a world where the answers lead to a wholehearted embrace of OpenAI.

We tend to underweight things that don’t happen. Every dollar LTF doesn’t spend is evidence of the political reality facing the AI industry.

The eventual data de-identification class action should at least save on outside counsel.

And a real insight into the experiences shaping the politics of millions of young Americans.

If you want an example of the kind of thing a Dem House majority looks into, it’s whether Nvidia’s $25M ballroom donation helps explain why Trump was briefed on a Chinese intelligence operation targeting Jensen Huang and did nothing.

Tying corruption/personal enrichment to rising costs is a potent attack. And it shows the danger of an industry getting too close to a POTUS ~25 points underwater. It’s an unpopularity doomloop. The tighter you hug, the more the hit sticks.

Comms-wise, the most interesting part of Anthropic’s S-1 is this line concluding the note from the 7 founders. It’s directionally spot-on. I’m also not at all convinced they’ve grappled with what earning that trust actually requires. To expand further: 1) They are correct that trust/credibility is a crucial driver of their bottom line. Recognition of that when attempting to make trillions off the largest economic/social transition in history is important, and not a given. 2) I think their chosen audiences are exactly right. Customers for revenue, governments for potential regulatory scrutiny/favor, employees to continue advancing the frontier, the public for some fraction of revenue, but mostly for their buy-in on a massive transition. I’d almost add in the larger “safety” community as well, as they will be the arbiters, so to speak, of whether Anthropic is actually making said decisions. 3) Recognition is not the same as action. And if they think their previous “public benefit-driven decisions” have succeeded in “trust-building”, they are delusional. Many of their enterprise customers are both concerned about control of their data and terrified Anthropic’s true aim is to displace them. Regardless of political belief, it feels fair to say Anthropic isn’t beloved by the current US Government. Employees, let’s call a wash since they clearly have significant talent, but also their most famous employee is one who quit for safety concerns. And the public at large, when it comes to opinions on AI generally, I mean yikes. If you can’t get that right when you’re a private company, it is not encouraging for how you will handle the stresses of being a public company. (FWIW, many of the same sentences apply to others at the frontier, which is part of the point – differentiating yourself on morality lines when you’re a trillion dollar + AI company is very difficult. But if your argument is that you will have business tailwinds due to that stance, it is certainly something you need to account for.) 4) The push-back would be that these decisions may have negative short-term impacts, but will be beneficial in the long-term. I think that’s a fundamental misread of the trade-offs inherent to these audiences, an underestimation of what @TheStalwart recently called the PMH (Politics Matter Hypothesis), and the reality that these choices will get harder, not easier, as the impacts of the AI transition are felt more acutely. 5) Re: the tradeoffs – these core audiences all want different things, and they are remarkably hard to please at once. As just one example: If you push the frontier faster to please your enterprise customers/accelerationists in Government, you scare the shit out of the public at large, piss off your employees, and animate the other forces in Government. If you slow the frontier down to stick by your principles and please your employees/safety voices in Government, you piss off your enterprise customers and the current Administration, without even the certainty of the public liking you more as a result. 6) So where does that leave you? Back at step one: convincing each of these audiences that Anthropic is best placed to be making these decisions on their behalf. Their core message basically boils down to: “Everyone going faster than us doesn’t care about safety, and everyone going slower than us has an inferior product.” If people don’t believe that, then the argument for those business tailwinds falls apart. Because you’re asking each of these audiences to trust your judgment – including when their interests are the ones you decide have to come second. That might feel like an exceptionally large burden – it is. It’s also the exact burden Anthropic has asked for. And to this point, they’ve done a lot more talking about what they aspire to be than convincing people to believe in their foundational value prop.

One of those times where you see the true value of “morally binding” accords. Trump and Silicon Valley set the terms. Then the lab expects the public to take its word for what happened. That’s two remarkably unpopular entities asking an awful lot of the American people.

To expand on the comms point — “this was all going to happen anyway” is just about the least compelling message. When you're chasing trillions and saying the downside was inevitable, don't expect gratitude. Expect people not to trust you to manage the risks.

Poll this tweet against your choice of Trump's AI message. It's not even close. We're in an exceptionally populist moment – anyone but Trump will struggle to capture that energy while still catering to industry. A real advantage for DeSantis/Hawley vs. Vance/Rubio ahead of '28.

Indeed. Just wait until January, when Senate Banking Committee Chair Elizabeth Warren starts asking why the current policy solution for agents is a bank's Terms of Service.

But don't people know how cute Muse is?

The political credibility Anthropic earned with Dems (in large part due to negative polarization from the Pentagon dispute) goes away quite quickly if they aren't seen as the reliable "safety-focused" company. Now, they might decide it makes the most business sense to just be lumped in with OpenAI (and public opinion wise, they might not have much of a choice). But if they truly believe that safety is their moat, they will have to do more work actually proving it.

I see this chart and all I can think is AI backlash. Access is a liability if you become synonymous with a system people think is rigged against them.

Look at Google/Spirit. Then read this. Add in Flock backlash & Americans' data getting sold to Chinese labs. Meanwhile every other tweet is about another billion for proprietary data. Put it in the column of accepted SV truths on a collision course with public opinion.