Last week, entrepreneur and philanthropist Joe Lonsdale announced a very intriguing initiative by the Abundance Institute, a pro-breakthrough tech think tank.

Here's how he put it:

"The Abundance Institute, in partnership with Stand Together, is raising $4 million to place a strike team of 15-20 AI-native software engineers, data scientists, and product leaders inside the FDA to accelerate the FDA's latest AI initiatives and bring outsider perspectives on new areas to iterate on fast. These leaders will remain Abundance employees, but under the Intergovernmental Personnel Act they can sit desk‑to‑desk with reviewers, wire modern data pipes into legacy silos, and automate the mind‑numbing paperwork that turns months into years."

It turns out that the announcement was premature since, after that, HHS put out a statement denying that this initiative was happening and saying "we have no plans" in this regard.

Since we were intrigued both by the concept and by the reaction, we reached out to Abundance Institute CEO Chris Koopman to learn more.

First, on the denial snafu. We understand that the announcement was a bit premature. Abundance Institute does not yet have the money raised for this initiative, and it is still in preliminary talks. However, Koopman told me, "We have, in fact, had encouraging conversations with the FDA. We do not have anything specific formalized with them, and we remain in active discussions." Obviously HHS was not going to confirm or endorse anything that wasn't already formalized and good to go, but neither is it just a white paper. Koopman and others are actively trying to make this a reality.

Now, what of the initiative itself? What would it consist of? What would it do?

If you talk to people who deal with the FDA, as we do, what they will tell you is that the most frustrating part of the process is not just the process itself, but the element of randomness, especially in the time it takes to get a response.

And this, Koopman told us, is not a coincidence, because the FDA uses antiquated work processes.

"Over the last several months, and speaking with a number of folks, former FDA officials, health care operators, innovators that are building next generation of tech enabled tools in the health space, what we kept hearing from people is that most of the bottlenecks aren't aren't statutory, but are operational."

As one hypothetical example of a project these engineers might work on, Koopman mentioned the adverse event reporting systems. "So when someone has an adverse reaction to to a pharmaceutical, to an OTC, whatever, it has to get reported to the FDA. This is how the FDA knows how to make recalls. And so right now that's done via emails and call centers. That's how the data is acquired. And then it's manually inputted across all of the centers within the FDA. And then there's a human interface of eyeballs on data, trying to trying to find signals among the noise." In other words, the decision whether or not to make a recall is based on someone emailing something, and then somebody else inputting that data in a form, and then other people looking at spreadsheets trying to figure out what's going on.

That seems like a really low-hanging fruit of something that could be expedited using AI and other modern workflows, by top-tier engineers.

As Koopman pointed out, "the FDA does not need an act of Congress to do this. The FDA doesn't need an executive order to do this. The FDA just needs programmers with the ability to build these systems for them, and walk alongside them to get them the updated tools and processes that they need."

So, how to fix this? Koopman again: "The opportunity we identified is that the current market value for any AI native engineer or any sort of programming talent that can operate on the modern tech stack prices the government out. Even for entry level roles in this space. The GS scale cannot operate in a world in which the tech talent market is as hot as it is. And so we thought: what if we could leverage the Intergovernmental Personnel Act?" More: "What if instead of trying to, like most nonprofits that do this, seek full or near full reimbursement from the government to cover the cost of these Fellows, what if we found enough philanthropic capital to support this?" In other words, "we wouldn't have to limit ourselves to whatever sort of structure there is in government so we can go attract and acquire top tier AI programmers, pay them something much closer to market rates than what the GS scale would allow, and then detail them into government. Basically, we absorb the costs and we just provide them as a service to support the the modernization efforts at the agency."

It seems like a very smart proposal to us. It seems like the kind of thing DOGE was meant to do. We hope it succeeds.

NOW: Subscribe to the PolicySphere Morning Briefing. Every morning, a briefing article and all the policy news you need to know.