Food for Thought: One-On-One with David Friedberg

I would say we're trying to just make better product for growers, and do it in a different way, in an advantaged way. So, you know, that's kind of how we think about it. We've developed a new system of breeding that we call boosted breeding, that really changes how we do breeding, the pace of gains, and also allows for true seed propagation in plants where we're typically doing vegetative propagation, where there isn't a seed business today. So it unlocks a significant yield gain and it creates new opportunities for creating seed products in markets where it doesn't exist. So it's innovative in that sense, it's new, it creates advantages, and, you know, in some sense it might alter some of the existing way that the industry operates, but we think it's ultimately going to be better for farmers, be more productive, etc.

So the way boosted breeding works is, traditional plant breeding happens through the sexual reproduction of the plants. The male sperm, the female egg combine, and the male sperm is half the genome of the father. That happens through meiosis. The two sets of chromosomes fuse, and so every sperm — and it fuses at different points, that's what makes every sperm different — and then the two sets of chromosomes in the mother fuse to make an egg, and it fuses at a different point, and that's what makes every egg different. So when the sperm and the egg come together, you end up back with two sets of chromosomes, and it's a random half of the mother and a random half of the father's genome. And so our system actually causes the sex cells to be produced through mitosis, meaning every sex cell has the complete copy of both sets of chromosomes in the mother plant and in the father plant. So when you combine them, you end up with four sets of chromosomes in the offspring rather than just two. So it's called doubling the ploidy — instead of being a diploid, it's a polyploid. And this system can be applied to a range of crops.

So part of our — and there are different advantages in different crops. For example, in potato, it will allow us for the first time ever to make genetically uniform true seed. So people have over the last 20, 25 years been trying to breed diploid potato — potato is a natural tetraploid — and they've been trying to breed diploid potato to inbreed them, cross them, and make traditional hybrid seed in potato. In our system, we can actually combine two kind of genetically distinct diploid potato to make a tetraploid offspring, but all the seed is tetraploid, so it's tetraploid true seed. And that creates a higher yielding potato, it allows you to stack traits, it allows you to have what we call more heterozygosity, more variation in the genome, and that heterozygosity results in a faster growing plant, more hybrid vigor, more yield, etc. So heterozygosity is the objective of boosted breeding, and making true seed is the outcome. So we can sell seed that is genetically uniform in crops that you can't sell today.

So in corn, which is published in our patent, you can see that our proposal is you can take kind of two corn hybrids, combine them, and you effectively end up with a tetraploid corn seed. And so in that case, you now have a corn plant that in our studies have shown is much, much higher yield than even kind of the best elite varieties that you can get in the market today. And so we are working in each of these different crop markets for different objectives. In corn, obviously, the objective is how do you really improve yield, how do you provide growers a significant yield advantage. In potato, there's a pretty straightforward opportunity: if you can make true seed, you can change the economics of potato farming. Potato farmers today use about 10% of their harvest to replant leftover potato vegetative biomass back in the ground — that's how they're doing planting. And if you've ever seen a video on this or visited a potato grower, it's an incredible operation. It takes a lot of fuel, a lot of labor, and it's pretty resource intensive. And then there's obviously a pretty significant fumigation effort that goes on. So when you add up the cost of the labor, of the fuel, of the resourcing needed to store all of what's called seed potato, all the fumigant, it can be as much as 40% of the revenue of that potato grower. And so we could reduce that cost down, we think, by over 75% if we can move to a true seed model. So those are the advantages in potato. Plus we can make better potato because we can now do breeding — the breeding system itself is better because it's predictable.

When you cross two plants together in a traditional breeding system, you don't know what you're going to get. You're going to get a random half of the genome of the mother and the father, so you end up with all this variation, and then you go and select through the variation. And yes, you can use things like genomic selection or marker-assisted breeding, but ultimately there's a variability that arises. In our system, when we combine two parent plants, we know exactly what we're going to get in the offspring. And so that changes actually how you can design a breeding system. It allows you to use quantitative genomics, which is software looking at the sequencing data of the parent plants much further up, and then make decisions about what do you want to combine together, what's the exact outcome you want. You can target the outcome, and then you can deliver the targeted outcome. So it changes the predictability of breeding. And so in some markets that actually creates a huge advantage where there's long breeding cycles. So there's a bunch of advantages to our boosted breeding system, both in terms of breeding timelines, in terms of heterozygosity and driving yield gain, and in terms of making true seed.

So to answer your question with my very long-winded answer, there's a lot of ways that we are looking at applying the system in different ways in different crop markets. And so we are working on a lot of different crop markets right now, all of which we work with in partnership with industry participants and industry players. And in that sense, we're really trying to make sure we're solving for the real need state and the real problems in the market, create real value, and have the right partners to help bring it to market.

Depends on the strength of our IP, which we feel is pretty strong. And I would say that, yeah, about a decade out, or let's say a few decades out — I mean, if you go back to, what is it, 1926, 1927, when George Shull hybridized corn and, you know, hybrid corn took off as an industry — we really do think that boosted breeding can usher in a new era of an alternative or a kind of scaled-up way of doing hybridization. And it allows you to stack together more distinct genomes into a single plant. And the benefit is — and there's a lot of studies on this, about three competing theories about why this is the case — but when you have multiple copies of a gene in a plant and you have different versions of each copy, the yield goes up, the vigor goes up, the growth rate goes up, the survivability goes up. Turns out plants are very good at accessing the gene they need when they need it. Their expression machines are constantly turning genes on and off based on the environment they're in. So the more unique copies of a gene we can put into a plant, the better they can adapt to their environment, which means the faster they grow, the bigger they get. So we think this is universally applicable, that over time we hope and expect to see that boosted breeding becomes almost like what hybrid breeding is today.

Yeah, I do. Yeah, I feel very confident in that. And just when you look at some of the results we've seen on yield gain, the yield gain we see is not the traditional kind of annual gain you would get with traditional breeding programs. We're talking about major step changes in yield and improvement. And then the speed of the breeding programs is dramatically altered. So, and then I will add on one additional layer: I think quantitative genomics is another big unlock. And I know a lot of seed companies that are running breeding programs, historically they started using kind of marker-assisted breeding, but now what we — the term du jour is AI. But at the end of the day, all of this AI, data science, big data, whatever you want to call it, going back 25, 30 years in Silicon Valley, there's always a name du jour. But the core concept is that as you get more data, you use statistics to make predictions from that data about the future.

A chatbot, an LLM, is the underlying large language model that drives a chatbot that we're all using that we call AI. That's actually a text predictor. They were able to suck up all the text on the internet and use that text to make a prediction on what the answer should be when someone asks a question, word by word. So based on the series of words that you put in your question, it gives you a series of words out, and that's a prediction engine. Prediction is now being applied to genomics using genomic data and phenotypic data. So as we collect a lot more phenotyping data in breeding programs and in field trials and in all of the machinery, farm equipment, satellite imagery that's collecting data now, and then we have all the genotyping data because it's so cheap to sequence stuff now with DNA sequencing, I can now build models that some people are calling genome language models, or, you know, whatever you want to call them, that can be predictive of a phenotype.

So I say, hey, I want 90-day corn, I want it to grow in this type of soil, I want it to have the following kind of characteristics, I want it to be drought resistant, deep roots — I can list all the phenotypes, and maybe the software can resolve the genome or the genes that would be ideal for that particular definition and for a unique environment, a unique soil type, this amount of sunlight, all of the components. And then if you've got different breeding systems, these advanced breeding systems, which some people call new genomic techniques, new breeding techniques, or, you know, we call it advanced breeding because we think it's inclusive of automation and all sorts of stuff that we use at Ohalo — you put all that together, and now that idea actually starts to become a little bit more of a reality. I can very quickly combine two sets of genomes where I know I've got the right targets in them, I can very predictably breed, I can even use some of these, you know, nuclease or, you know, tools that can go in and make specific alterations to specific genes that I know are natural genes native to the plant, that if I make this particular change it'll drive this phenotype. So when you combine all of the tooling with the predictive capability, suddenly quantitative genomics becomes a real reality.

And so you put all of this together, and I feel very confident that, you know, a lot of the big-picture objectives are not just achievable, but I think we're going to have a lot of abundance. And I think we're going to have a lot of questions in the next decade or two about acres switching over, because suddenly acres — and a lot of people listening to this might be thinking about corn and beans in the US maybe, but I really think around the world and just generally speaking, we're going to have seed that will allow greater productivity in a particular acre and a geography that's limited today in terms of what's available to them from a crop perspective. So anyway, I think a lot's going to change.

Yeah, because of the productivity.

Exactly. So it feels like we're just on the cusp. These things are now achievable. The technology exists.

Well, it depends on how things adapt. You know, there's a lot of this that we see technology typically only being rate-limited by social behavior, and this is true in consumer and in business. In business, I think we will see different rates of adoption, different rates of embracing different technologies, different tools, and that'll ultimately be kind of the true rate limiter on the utilization of these technologies and the output of these technologies. So I can only be so predictive. Anytime I've tried to predict social behavior, I've been generally wrong.

Me and everyone. So it's a very kind of gray space. I mean, look, GMO technology is a great example. No one really knew what was going to happen when it started to happen, that people freaked out about GMO technology, saying it was this boogeyman, bad thing, and it limited the utilization and expansion of GMO technology, which could have had a pretty profound impact on the world. We're in a new era, I think, right now. The regulatory environment on gene editing is correct, it is favorable, it is being deregulated, and that is a positive thing. But there are other aspects of adoption that I think matter. Farmers that have been farming potato one way for 30, 40, 50, 60 years are not going to switch overnight to using a true seed system and change out their entire operation. They're going to want to wait to see their neighbor do it for a couple years. And, you know, when you play that out, the question is who's going to move first, how are they going to move, how are the changes going to be embraced.

And, you know, I think one of the things we're all going to be surprised by is that there will be a pretty rapid pace of adoption in emerging markets for some of the new technologies that we're bringing out, faster than developed markets. And that's because I think the benefit, the economic benefit to the smallholder farmer in emerging markets is so profound, it will actually drive a much more accelerated, widespread adoption than we'll see in an industrialized setting where you have too much to lose and not enough concept of what the upside is. Whereas in emerging markets, you have very little to lose and a ton of upside. And so I'm actually pretty encouraged by some of the work we're doing in emerging markets, where I think, you know, we could see much faster kind of pace of adoption. The trade-off is for our market here, we risk being back on our heels if other markets beat us to it in the developed market.