Jack is the CEO of Hidri SandboxAQ,Thank you for reading this post, don't forget to subscribe!
Artificial intelligence is advancing with new tools that take us far beyond writing essays and summarizing documents. While large language models (LLMs) are adept at processing text and visual learning models (VLMs) are excellent at creating graphics and short videos, another set of AI tools are needed to address the most pressing challenges of our time: diseases. developing medicines for, accelerating the move toward renewable energy and storage, creating new materials for automotive and aerospace, developing safer and more efficient water filtration, and tackling other key goals.
This is where AI simulation steps in. AI simulation combines the techniques of quantum physics with the power of deep learning to address these challenges.
Let’s focus first on biopharma development. McKinsey published research showing that “the average time for a new drug from candidate nomination to launch has been approximately 12 years.” As far as cost is concerned, a report by Deloitte found that “the average cost of developing a new drug increased by $298 million to $2.3 billion in 2022.” We need cost-effective and timely drug development that produces safe medicines. The sooner we have these, the faster we can make a positive impact on patients’ lives.
Other important issues we must solve include introducing new battery chemistries to power EVs and storing renewable energy on a large scale. We also need a way to remove chemicals (PFAS) forever from water filtration and other key areas of our lives. To solve these big problems we need to go beyond just text or image-based AI.
The AI simulation revolution has just begun
Artificial intelligence models like ChatGPT perform well when they are trained on large amounts of data. However, when there is not much data – which is the reality in most cases in the physical world, what do we do? For example, if we had a lot of data on a new drug for Alzheimer’s disease, we would already have solved the problem. There is a need for a new form of AI that can generate high quality data based on real-world physics and dynamics. This is where simulation comes into the picture. A simulator is a computer program that contains the key equations that govern molecular or other dynamics and can run possible combinations and interactions billions of times. Deep learning AI is applied to this data to optimize for the desired outcome, such as a drug with high binding affinity for the target receptor.
Running billions of simulations of a molecule’s digital twin in a large GPU cluster computer model is far easier, faster, and less expensive than manipulating a physical experiment in a laboratory. Additionally, AI models can learn from previous simulations to make further improvements.
Consider Complex, a model created by researchers at MIT and Tufts University MIT News, enabling them to “screen more than 100 million compounds in a single day”. Simulation is especially helpful for driving progress in challenging, unavoidable conditions like cancer or Alzheimer’s, where little data exists for AI models. In this case, quantum mechanical interactions between drug compounds and human receptors – simulated on today’s classical computing hardware (GPUs) – create new data from physics and unlock insights that have eluded researchers for decades.
Using simulation, researchers can create drugs that are safer for patients and have a better chance of succeeding in clinical trials by running virtual scenarios of drug interactions within the human body before the first person takes the drug. And they can do it faster and more cost-effectively by eliminating years of laborious laboratory work trying to replicate molecular-level interactions to find the most promising compounds. This ability to influence the physical world is why some GPU makers are doubling down on these new forms of AI.
Other major use cases of AI and simulation
The combination of AI and simulation goes beyond medicine. The combination of both technologies helps scientists and others make cutting-edge discoveries in various fields with less time, money and risk than traditional methods.
Clean energy is another major use case. While lithium-ion batteries have been integral in helping to power electric vehicles and store solar and wind energy, they have major limitations. We need batteries that have higher energy density, weigh less and are less expensive to produce. However, performance-testing of advanced battery chemistry, materials and designs can take years through traditional methods. Using simulations, researchers can explore countless combinations using AI to optimize designs and make performance predictions much faster, helping to bring these new batteries to market or optimize them for specific applications. Is available.
Other major use cases include materials science, food technology, and cosmetics product development. If it is a tangible object, there is a possibility that AI and simulation can work together to help researchers create a better version of it. For example, using these two technologies, scientists can create new types of biodegradable plastics or more sustainable building materials, reduce harmful additives in the foods we consume, and improve skin care and beauty products. Can create products that do not have some of the toxicities we see today.
real world testing
AI simulation opens up many new possibilities for tackling big problems. To validate the results of these computer models, we still need to put the final suggested molecular structures through real-world testing. There is a constant feedback loop between AI simulations and physical testing that drives progress. This is important not only for security but also for improving the AI models in use; The more reinforcement learning researchers incorporate into the development process, the faster models will improve.
Now by strategically using AI and simulation together, we can make breakthroughs that will create commercial and societal benefits. Key to this success will be training engineers and leaders in these innovative AI tools. Universities, corporate and government leaders all have a role to play in driving this skills upgrading forward. Let’s make sure these powerful tools are democratized around the world so everyone can benefit from them.
The Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs, and technology executives. Am I eligible?