The Paul G. Allen estate has rolled out its latest philanthropic effort: the Fund for Science and Technology (FFST). Backed with a US$3.1 billion endowment, the FFST plans to invest at least $500 million over the next 4 years into bioscience, artificial intelligence (AI) and environmental research.

Credit: Fund for Science and Technology

“Our commitment is on big bets, long bets, that can solve really, really hard problems,” says FFST CEO Lynda Stuart. Stuart has helped place similar wagers before, working for example from 2013 to 2022 at the Gates Foundation, overseeing the development and distribution of vaccines, biologics and antibody therapies. Following a brief stint on vaccine development at BioNTech, she then served as Executive Director of the Institute for Protein Design (IPD) from 2023 to 2025, during which time the IPD’s founder David Baker won a Nobel Prize for his contributions to computational protein design.

With the FFST’s first tranche of bioscience funding, Stuart and her team are supporting immunology research at Benaroya Research Institute, cell therapy work at Seattle Children’s and both cancer immunology and cell therapy at the Fred Hutchinson Cancer Center. These grants could help the community to figure out how to engineer cells to treat diseases, she adds. Future investments will broaden the fund’s scope further still.

“There are a lot of amazing, really hard problems that we need to tackle as humanity,” says Stuart.

What’s your vision for the FFST?

I was quietly sitting in my academic role at IPD, and I had this mysterious phone call about this new foundation that wanted to support three of my favourite things: bioscience, the environment and artificial intelligence for good. That is a very compelling set of things. And we’re really interested not just in funding within a pillar, but funding at the intersections between those pillars. That could allow us to tap the transformational power of not one but maybe two or three different technologies at any one time.

What’s amazing about this opportunity is its real commitment to impact, a real ability to move and work at scale. Being able to do my small part in forwarding [Microsoft co-founder] Paul Allen’s visionary legacy is really an opportunity that was impossible to turn down.

Another thing that is really exciting about the FFST is that we are able to reimagine how we think about funding science, how science is done, how collaborations come together, and how we can break through boundaries that have plagued mankind for some time.

We want to make sure people can continue to dream and imagine really different futures. That’s who, why and what we are.

How will you re-imagine the funding of science?

A lot of funders want proof of concept. One thing that we can do is take big bets, very early on, on something that might just be an idea. And then we are willing to follow that idea for long enough to see it have an impact in the world. FFST is very committed to funding across that whole value chain: we’re able to provide grant funding at the beginning, and if we’re in product development mode we’re able to leverage programme-related investments to acquire equity in initiatives with the potential to spin out.

There are lots of tools in the philanthropic toolkit, and we are willing to use pretty much all of them. That enables us to take an idea from ideation the whole way through to impact.

You’ve described FFST as a ‘discovery organization’. What does that mean for you?

There are three things I would say that make us a discovery organization. One, we’re willing to fund things that are just knowledge, just learning about something new. Like mapping the human brain, which is what the Allen Institute is working on. Two, we can fund the development of tools that are field-enabling. And three, we can fund solutions.

In the context of drug development, that might mean taking something from an idea through to preclinical proof of concept — through that translational valley of death. But we wouldn’t be doing phase II or III trials that other funders might do.

Our area of expertise will be that earlier riskier part of the pipeline.

How do we see that in FFST’s first bioscience grants?

Backing up a little, we’re from Seattle and so we wanted these first investments to be in Seattle, supporting these real centres of excellence here. We really wanted to shine a light on these extraordinary organizations.

One of these is the Benaroya Research Institute, primarily an immunology institute. What we’re supporting there is their phenomenal biorepositories, which have both longitudinal and cross-sectional samples from people with a bunch of different autoimmune diseases including diabetes, rheumatoid arthritis and colitis. Those biorepositories provide this foundational resource for so many other questions and opportunities, including target identification and hypothesis generation. Benaroya has been collecting them for 20 years, and they could never be replaced. We want them to continue to collect them for another 20 years, and we feel that they’re just a great base for all the work we and others might do in these fields. That’s a foundational investment.

A second grant was to Seattle Children’s, which has a cell therapy programme that is trying to bring cell therapy into paediatric orphan disease populations. That’s an extremely important area, but it’s also a bit of a neglected area. It’s often small populations, not big commercial markets, and you can’t do classical clinical trials in these diseases. We’re really interested in helping them reimagine how to bring those interventions to patients in ways that don’t really fit within the big pharmaceutical models. We’re just at the beginning of a very interesting conversation here, and I don’t know how it will play out.

And then the Fred Hutch Cancer Center, an amazing cancer centre, is really interested in both cell therapy and tissue immunology. We are supporting their work in trying to understand how the immune system works in different contexts.

One theme that comes out of all these is how do you engineer cells to do magical things?

We hope that by funding these three organizations, we can also bring them together to really foster cross-collaborative efforts that think differently about those problems.

Will future grants fall under that same theme?

I’m quite interested in engineering cells to do magical things, and I think that might be a theme in our long-term strategy. But there’ll be other themes for sure.

Cell therapy is currently going through a rough patch, with several large pharmas, biotechs and venture capitalist groups pulling back in this space. Did that drive your decision to invest here?

I wouldn’t say that motivated us specifically. And it is often the case that the first wave of a field will hit a wall. When that happens, you need to retreat back into discovery and rethink things, and then you can break through again. There are a number of fields that have gone through that paradigm, and that’s something we’re very comfortable taking a risk on. The concept still makes a lot of sense, but there’s still science and discovery to be done. We’re happy to fund the tinkering that is required to optimize those sorts of things, so that in the long term we can get to the upside.

I think maybe it’s a good time for cell therapy to fall back to philanthropic funders, and it can move back to commercial when it’s ready.

You’ve not yet disclosed any AI investments. What’s on the horizon there?

AI in the life sciences and in bio-engineering is clearly an emergent field. David Baker and colleagues at the IPD have done a lot of work in proteins, but there’s similar work that needs to be applied to DNA engineering, RNA folding, enhancers, repressors, and synthetic pathways. Those are types of areas where AI will be game changing.

We haven’t made these commitments yet, but I can imagine there being a lot of opportunities for AI also in diagnostics, tissue imaging and cell imaging. These are areas that are ripe for AI.

One of the things that we want to do as we develop those AI capabilities is to do so responsibly. Where appropriate, we will want to make these open. If that’s not appropriate, we would want to at least make sure the tools are accessible to non-profits that are working in that space. We don’t want these capabilities to get locked up only at for-profit places.

There’s a lot of hype around AI in drug discovery, but few clear cases where it has as yet enabled truly better, different or faster drug candidates. Given your prior role at IPD, what are you watching for near-term wins?

Backing up even further, when I was at the Gates Foundation one of the things that I became quite obsessed with was how to manufacture things at scale cheaply. There are a number of programmes with great potential that I’ve seen, but that aren’t manufacturable. GLP-1s are actually a good, recent example of something where the manufacturing is a real bottleneck.

If you go back into the history of those products, someone often needs to make a decision in discovery to re-engineer the product to enhance its manufacturability. But people don’t want to do that, because that could set them back 6 months to a year. And then 5 years down the line, they can’t fix it at all because they’re locked into a product. Fast forwarding, where I think AI could be transformational is if people use it to think about designing products for manufacturability right from the beginning. Researchers can design drugs for production at scale, and AI tools can be used to maintain the functionality of the molecule.

That’s where I would love to see AI being used more: not just in terms of what you discover, but also how you get it through that valley of death and out on the other side. I think that’s a real untapped potential.

You recently co-wrote a paper on AI for ‘undruggable targets’. What do you see as the low-hanging fruit there?

I’ve been really fascinated by the opportunity in intrinsically disordered regions (IDRs). Something like 40% of human proteins have IDRs in them. These IDRs are functional and important, but we don’t even know what they do because they aren’t druggable. Sometimes with a field like that, you just need a hook to get started. I’m hoping that some of the work that the IPD was doing around IDRs will provide us with the hook that will open up that space.

We’re not there yet. These IDRs are largely in intracellular targets, and so biologics are not really good at taking them on. But I think this is a space where AI might provide a gateway to understanding new drug target classes and other therapies. It’s super interesting.

Stepping back to your prior role in vaccine development, what do you make of the CDC and the NIH’s moves away from vaccine discovery, development and access?

Clearly I believe in vaccines. I trust the data on both their safety and their efficacy. And I continue to believe in their importance. I come from a developing country, and have seen people get sick because they didn’t have access to vaccines for preventable diseases. I continue to believe that they save lives in an amazingly transformational way.

When I step way back and look at what’s happening, I think it’s really important for us all to understand and trust in science. How do you introduce an innovation into the world, and how do you make sure that that innovation is accepted? That’s something that FFST is really committed to as well. As we develop and deploy new innovations, we will think about this from the get go so that they have good acceptance. I think maybe that’s what we’re seeing here: by necessity COVID vaccine development had to move very fast, and that created some concerns for people.

I’m hoping that the world’s rationality will come back, and we will focus again on the really positive impact that vaccines can have.

And what about the huge funding cuts at the NIH? Can philanthropies fill that gap?

Philanthropy can’t fill the gap of the federal government. What we can do is be much more strategic about how and what we fund. I think we can have a lot of impact, and our commitment at FFST is really to be smart with how we fund things so that we maximize that return. But nobody can step into the federal gap.