Imagine if Artificial Intelligence (AI) could be used to reduce research and discovery cycles of new ingredients by years at a time, saving millions of dollars. Or if AI could create a tighter more cost-efficient supply chain or produce new, differentiated products in a more effective way.

Imagine if AI could reduce waste in all aspects of the business to improve overall margins. Then consider the numerous ways that AI could empower employees, for example, making it possible to unlock consumer trends months ahead of the competition and subsequently address those trends with timely, cost-effective marketing campaigns to precisely activate the right consumers at the right time with the right solution.

For companies in the food supplement, food, and medical food space who are open to innovation, AI can help enhance the employee experience and the consumer wellness journey, thereby creating growth for companies. The examples imagined above are just some aspects of what AI systems can do when effectively trained and managed.

Artificial Intelligence (AI) encompasses a broad field of computer science dedicated to creating systems capable of performing tasks that typically require human intelligence. AI is data driven, something which appeals to me as a mathematician with a deep devotion to the intense complexities of human biology. When I founded Nuritas ten years ago, it was my background that first compelled me to embrace the potential of AI. Today, Nuritas specializes in discovering new groundbreaking peptide ingredients for the industry.

Given our company’s deep expertise in AI, it is no surprise that I am often asked about what AI is, how it can help advance our industry, and how we can use it in our daily business lives. Let’s start by reviewing some of the basic concepts of AI.

AI is often classified in two main categories known as narrow (or weak) AI and general (or strong) AI. Weak AI refers to AI systems that focus on specific tasks, such as image or speech recognition, natural language processing, or playing chess, but that lack general cognitive abilities like human intelligence. Strong AI seeks to replicate human cognition and intelligence, such as comprehension, learning, and reasoning. However, achieving human-like cognition and consciousness in machines currently remains more of a philosophical concept than a technological reality.

Machine Learning (ML), a subset of AI, enables computers to learn and make predictions from data without explicit programming, focusing on algorithms and statistical models. Generative models, in contrast to discriminative models, are designed to create new data that resembles the training data that the AI system learned from.

Deep learning (DL) is a subcategory of ML using multi-layered, neural networks that learn complex data representations. For example, deep learning models like AlphaFold or ESMFold have significantly advanced protein structure prediction, surpassing traditional methods in accuracy and speed. AI's role in protein folding is crucial in biology, predicting the three-dimensional structure of proteins from amino acid sequences to understand their functions in cells. DL has revolutionized AI, driving breakthroughs in areas such as robotics, for example.

Think of an AI system as you would a baby, one that is dependent on being taught. To harness the power of AI, you need to responsibly feed that baby the right information to ensure that the baby will learn. You need to rely on knowledgeable experts to teach the baby what it needs to know.

For example, the outcome of your experience with AI is going to be dependent on how well you teach it to make the correct decision from an incorrect or insufficient decision. This will depend directly on the type of data you give it which is why I always encourage all companies to start storing their data in a format that is not only AI friendly, but also one that allows decision making. This will cost the company very little relative to what it could use the data for in the future. Following these procedures will also allow companies to learn from all their past mistakes and see new interpretations and trends they would have otherwise missed.

Earlier in this article, I asked you to imagine some of the ways that AI can improve your business. Most of those examples have very finite and usually well-understood parameters. As such, creating training sets for an effective AI performance in those cases is not as difficult as is reducing the discovery cycles of new ingredients. That is the hardest of the examples to manage, because biology is mathematically the most complex entity known to humans, leaving us with many gaps that have not yet been filled. Harder, yes, but not impossible, as that’s what we are doing at Nuritas.

Let’s now take the theoretical value of AI and bring it to a practical level, using Nuritas’ experience in taking a molecule from concept to market with PeptiStrong, our patented, branded ingredient, a hydrolysate from fava bean protein and a network of other peptides that support muscle health. Using our own AI platform (1) we discovered a peptide molecule that otherwise would have remained hidden.

Peptides are short chains of amino acids, found in nature and in our bodies. We need them for proper cell function, to ease inflammation, to turn food into energy, and more. Our AI platform is trained to identify peptides in nature, from plant or marine sources, as examples. And not only does it find these peptides, but it determines which ones are practical to bring to market: are they reproduceable in appropriate quantities; will they survive, for example, the journey through your gut, your body’s temperature or pH levels; will the molecule be bioavailable and stable through the extraction and manufacturing process?

AI can look at multidimensional criteria that we as humans can't, and then it narrows down the space of peptide molecules very quickly to the few that have an activity. In this way, AI takes the guesswork out of identifying which peptide complexes to pursue, not only speeding up the discovery time of a new peptide molecule, but also creating cost savings in the process.

From there, you still need to back your claims with science when moving from concept to market. This gives customers a way to differentiate their products and ultimately provides consumers with specific health benefits. Using PeptiStrong as the example, we’ve done just that. With three completed human clinical trials, we are confident that a 2.4 g dose supports these clinical benefits: increases muscle protein synthesis, reduces muscle breakdown and exercise-induced inflammation and lowers myostatin while increasing energy (2) (3). The newest study (4) not only shows impressive results for increasing upper and lower body strength compared to placebo, but also a significant increase in endurance in just four weeks, as well as a 1% increase in overall bone density after just 56 days.

So, what AI brings to the table for the food supplement, food, and medical food industry is not only an improved discovery time for new ingredients; it also allows for the discovery of efficacious, scientifically proven ingredients, something that has been impossible in the past without years and years of R&D. Besides this, AI also can help us find new ways to understand our own biology and create clearer cut, less random, nutritional ingredient solutions that will revolutionize the way that we, as an industry, will look at how we identify ingredients that are more adaptable, more accessible, and less wasteful.

Particularly through generative models and advanced ML techniques, AI can transform scientific research and industry practices, driving innovation in healthcare, and beyond.

My team and I are immersed with meeting manufacturers and marketers in the food supplement, food, and medical food space, as we seek to be their innovation partners for current and next generation healthy products. What encourages me is that rather than fear of AI, many of these companies are intrigued, understanding that AI is the future in this space.

There is less resistance and more interest in practical business questions about AI. Questions such as: what can we do with you with this new technology; how do we know it works; is it better than what exists; how do these ingredients discovered by AI perform; and how can I use that ingredient within my finished product?

These are all reasonable questions and they’re the same kinds of questions that you’d get about any new technology hitting the market.

AI techniques are making substantial impacts across practical applications. In drug discovery and computational chemistry, AI accelerates identifying new drug candidates by predicting molecular properties, screening compound libraries, and optimizing drug designs. In computational biology, AI analyzes complex biological data like genomic sequences and protein structures, yielding insights into disease mechanisms and personalized medicine.

I’m convinced that AI will also strongly take hold in the nutritional space, so much so that in the future we’ll forget how we used to find ingredients, as a properly trained AI platform will prove better results than what our brains could do on their own. AI is an enhancement tool that focuses our thoughts on the areas that need our input and less so on just random problem solving. It elevates the way we think and is a wonderful partner for our brains and a strong ally for our industry.

Obviously, implementing lifestyle as medicine at a large scale involves some hurdles. To overcome them, we will require innovation, creativity and ecosystem-thinking, combined with biomedical and digital technology and novel solutions from agri- and food. Moreover, new governmental policies should support such changes in order to sustain them.

One of the biggest challenges is the achievement of a long-term lifestyle behaviour change in people with a chronic disease. Here, personalisation comes into play. Ideally, biological cause as well as socioeconomic and cultural background, personality, preferences and goals should be considered in the treatment. In type 2 diabetes, for example, it is known that different subtypes of the disease exist with a different biological cause, etiology and complications, requiring a tailored treatment strategy. Furthermore, it is pivotal to make the invisible visible, especially to show the impact of the lifestyle treatment and a person’s behaviour on the biological cause of the disease. In particular, continuous monitoring coupled with just-in-time adaptive interventions have shown to be very powerful in lifestyle behaviour change and in stimulating people to regain to healthy habits in case of a relapse. Blended solutions, in which digital health innovations are combined with face-to-face contact from healthcare professionals, lifestyle coaches, peers, peer coaches or buddies showed to be scalable and cost-effective for lifestyle medicine.

The ‘magic pill’ for chronic, lifestyle-related diseases is personalised lifestyle medicine. For this, we need all of you! To achieve a healthy society in which we live long and in good health, we need to collaborate and combine forces to create fun, tasty, convenient and easy-to- adopt agri and food and high-tech solutions. If we work together, a very large market will welcome us and more importantly we can help to reduce the chronic disease burden and to maintain a high quality health care system.