Neal’s word. Synthetic. And generative.

by Neal Gutterson

The ag crop protection industry is a roughly $80B global segment, with major pressures on current products and how they are applied coming from consumers and regulators, and with biological realities leading to the emergence of resistance and technological innovation that hasn’t kept up. There is a significant investment in disruptive technology and business models, from new categories of biologicals to precision applications. Much of the disruptive innovation investment comes from start-ups, with major investments in biological alternatives to synthetic crop protection chemistry.

However, the demonstrated market opportunity for new and improved synthetic chemical crop protection products still dwarfs the market for biological products. While the market is visibly pulling us towards more biological solutions, most farmers know that their bread-and-butter solutions for protecting their crops come through synthetic chemistries. At Radicle Growth, we continue looking for great investments in biologicals while recognizing that we need new innovations insynthetic chemistry.

So, is this a good time to invest in synthetic crop protection chemistry innovation? We believe the answer is yes, but there are not that many start-up companies in this segment. Also, with the power of incumbents and the cost to scale innovation without partners, exits can be challenging. So, we approach this investment opportunity with enthusiasm but also with caution.

Other encouraging signs for synthetic chemistry innovation have appeared. The European Union recently backed away from its commitment to reduce pesticide load [by] 50% by 2030, thanks to farmers flexing their muscles across the continent. Overall, there is more reason than ever to have conviction about investing in technologies that can transform synthetic chemistry discovery and enable the development of new modes of action products with improved regulatory profiles.

But knowing that a process needs transformation is insufficient, of course—we also need conviction that new technologies are available or close to it. Fortunately, this is an exciting time for synthetic chemistry innovation, with new AI tools, better computational modeling (the advent of alpha-fold, for example), and deeper biological insights all coming together at the right time.

Of course, we’ve seen technological waves come and go despite exciting science. Often, there is a disconnect between real pain points and perceived pain points or about the most challenging aspects of chemical discovery, so investors sometimes invest in solving the wrong problems.

Nonetheless, the time is ripe for innovation in synthetic crop protection discovery. And this brings me to my second keyword, generative. We are exposed to tremendous hype about AI and what it can do for us. How can AI-informed decision-making improve farming for example? Can we improve farming systems overall through computational modelling? The jury is out on how to capture value from customer-facing uses of AI. However, the creation of new seed and crop protection products is being realized now and will be even more in the years ahead.

In the earlier days of machine learning, from the 1990s to the 2010s, AI was focused on pattern recognition, defining the features that enable highly effective categorization. Is this a cat or a dog in a picture? Is this cancer or not in a diagnostic image? Is this a weed or not? However, with the advent of generative adversarial networks (GANs) in the mid-2010s, we shifted from identifying the features that distinguish categories to exploring feature space to create better-performing examples of those categories.

In short, using generative methods, we can move from identifying to creating to developing novel chemistry. My first exposure to GANs was in breeding applications, where the amount of data on varieties performing better or worse than a reference was enormous, and haplotypes were readily mapped. As a result, breeders could generate haplotypes computationallywith predicted superior performance.

As haplotype space can be explored to predict better-performing crops, chemical-feature space can be explored to predict better-performing small molecules. A lot more data is needed to match better and worse chemical performance within the enormous feature space of chemicals for a range of different indications. Further tool development is also needed. With sizeable investment in the pharmaceutical industry for new drug discovery, we can expect spill-over benefits for the crop protection industry.

Between the massive market opportunity in synthetic chemistries to protect our food system and new, game-changing AI technologies, the future is bright for our farmers and disruptive innovations in synthetic AND biological crop protection products.



This entry was posted in Radicle Press. Bookmark the permalink.

Comments are closed.