A viral X essay reached over 30 million readers with a message that should unsettle every think tank leader in the country: “I am no longer needed for the actual technical work of my job.” The author, an AI startup founder, described watching artificial intelligence go from useful tool to full replacement for his core professional output in a matter of months.
Commentary like this is everywhere now. But almost none of it is directed at think tanks yet.
Think tanks are cognitive-work organizations to their core. Every function that defined the traditional model, research, policy analysis, communications, written testimony, public comments, is precisely the kind of work AI is rapidly learning to do.
This matters urgently for leaders in the State Policy Network, not because think tanks are doomed, but because the competitive landscape for policy influence is being reshaped in ways that demand strategic attention right now.
Three Forces Are Converging
First, AI creates an attention paradox. It makes your organization more productive, but it does the same for every competitor. Content supply explodes while the number of policymakers remains constitutionally fixed. There are no more state legislators today than in 1980. The result: Content abundance creates attention scarcity, and the historical value of being the organization that produces quality policy analysis erodes as everyone gains that capability.
Second, the content flood creates a trust premium. When policymakers are drowning in AI-generated material, they default to trusting people and brands they know. The government affairs director with seven years of relationship with a committee chair doesn’t become less valuable in this environment. She becomes irreplaceable.
Third, AI enables what I call the “individual bypass.” A single talented person with AI tools now operates the equivalent of a research department and communications shop with zero employees. Social media and Substack give them distribution. A large enough audience gives them influence with policymakers. These individual superstars will continue to rival established think tanks for attention.
Here’s the Good News for This Network
The organizations that have built Durable Freedom Infrastructure over the past decade have already made the essential strategic bet, even if they didn’t frame it in AI terms. The DFI model was built on the insight that writing papers wasn’t enough to win. That insight is about to become dramatically more true.
AI raises the floor of what we can do. The tech is only as good as the inputs and our feedback. Working with AI seems to be the sweet spot.
What used to take days and dozens of man hours can now be a starting point. Content creation, for instance — whether it’s research, writing, editing, newsletters, donor emails — should all get better, more targeted, more compelling, and more clear. Plans should get more specific. Strategies should get more sophisticated. And our capacity for impact should grow.
Consider what AI cannot easily replicate. It cannot replicate the litigation capacity that requires institutional legal resources, bar admissions, and years of courtroom credibility. It cannot replicate the grassroots network that fills a hearing room with real constituents on 48 hours’ notice. It cannot replicate the community engagement leader who has helped families navigate school choice enrollment for years and knows the local school board members by name.
It cannot replicate decades of accumulated knowledge about your state’s political dynamics, the kind of knowledge that tells you which committee chair is persuadable and exactly how to frame the argument. And it cannot replicate the genuinely creative policy insight, the novel solution to a problem that doesn’t yet appear on either side’s policy menu.
These are the capabilities that become more valuable, not less, as AI floods the zone with cheap content. Relationship depth, execution capacity, local knowledge, and true policy innovation are moats that neither artificial intelligence nor a talented individual with a Substack can cross.
What Should Think Tank Leaders Do?
Start by honestly auditing how your staff spends its time. What percentage goes to content production versus relationship building versus direct execution? If AI could handle much of the content production tomorrow, would your organization still be essential to your state’s policy landscape? If the answer isn’t an emphatic yes, your strategic priorities need to shift.
Experiment with AI tools now. The window where early adoption creates competitive advantage is open but closing. Give your team access to the best tools and a mandate to experiment and attack bottlenecks, not to cut staff, but to free capacity for the work AI can’t do: deepening policymaker relationships, building community presence, expanding litigation and political capacity, and showing up in the places where physical presence and human trust are what matter.
Rethink your next five hires. You may need fewer excellent writers and more excellent relationship builders, organizers, litigators, and political operatives, people who use AI to handle content production while they focus on work that requires human judgment and presence.
The think tanks in this network that have invested in DFI, that have built litigation centers and grassroots networks, that have embedded themselves in their communities and their state capitols, are better positioned for the AI era than they may realize.
The organizations that win in Washington State and Oklahoma and Pennsylvania didn’t win because they wrote the best white papers. They won because they were impossible to ignore: present in the courts, present in the communities, present in the capitol, present at the ballot box.
AI doesn’t change what it takes to win. It clarifies it. The think tanks that last won’t be the ones that produced the most content. They’ll be the ones that were impossible to replace.
