He’s probably wrong about the drugs.
Ray Kurzweil writes:
Having worked in the field for 61 years, longer than anyone alive, I’m pleased that AI has been at the center of the global conversation. But most commentary misses how large-scale language models like Chatgpt and Gemini fit into a much larger story. AI is poised to leapfrog beyond just revolutionizing the digital world to transforming the physical world as well. This will bring countless benefits, but three sectors in particular will be impacted most significantly: energy, manufacturing, and healthcare.
This is an excerpt from “Ray Kurzweil on how AI will change the physical world.” economistJune 17, 2024 (Deadline)
Kurzweil makes his point well.
Another excerpt:
In contrast, AI can quickly sift through billions of chemicals in simulations, and is already driving innovation in both solar power and batteries. This is poised to accelerate dramatically. In history up until November 2023, humans had discovered about 20,000 stable inorganic compounds that could be used in any technology. Then Google’s gnome AI discovered far more, and that number rose to 421,000 overnight. But this is just a small part of the applications of materials science. Once a much smarter AGI (artificial general intelligence) finds the perfect material, large-scale solar power projects will become viable, and solar energy will be abundant enough to be available for almost free.
Energy abundance will enable another revolution in manufacturing. The cost of almost all goods, from food and clothing to electronics and automobiles, stems primarily from several common factors: energy, labor (including cognitive labor such as research and development and design), and raw materials. AI is set to significantly reduce all these costs.
Where he falls short is on pharmaceuticals. It’s not that he doesn’t make a compelling case that in a relatively unregulated market, AI could easily have a huge positive effect on the types of drugs we put in our bodies. The problem is that he seems unaware of the enormous power the Food and Drug Administration has over what drugs we can take.
He writes:
Although much more laboratory work is needed to accurately simulate larger scale simulations, the roadmap is clear: Next, AI will simulate protein complexes, then organelles, cells, tissues, organs, and eventually the entire body.
This will ultimately replace current clinical trials, which are expensive, risky, time-consuming and statistically underpowered. Even in Phase 3 trials, there will likely not be a single subject who matches you on all relevant factors: genetics, lifestyle, comorbidities, drug interactions, disease variability, etc.
Digital trials will enable personalized medicines, with staggering potential to treat diseases like cancer and Alzheimer’s, as well as the detrimental effects of aging itself.
This will only happen if the FDA effectively backs down. Let’s hope, but hope should not ignore the painful lessons learned from experience.
