At the recent World Economic Forum held in Davos, Demis Hassabis, CEO of Google's DeepMind, stated that the first drugs designed with the help of artificial intelligence could begin clinical trials by 2025. Hassabis is also the head of Isomorphic Labs, a drug development company under DeepMind. He mentioned, "Our plan is to have some AI-designed drugs enter clinical trials by the end of this year."
Image Source Note: Image generated by AI, licensed by Midjourney
Since 2021, Isomorphic Labs has been dedicated to using machine learning to accelerate drug development. Hassabis noted that personalized medicine is expected to be realized in the future, with AI systems capable of optimizing drugs based on each individual's metabolic characteristics in a short time. He emphasized that pharmaceutical companies are increasingly interested in AI, as it has the potential to save them significant time and money.
According to an article in Nature Medicine, the process of developing and approving new drugs typically takes 12 to 15 years and costs up to $2.6 billion. Moreover, over 90% of clinical trials fail. Therefore, any technology that can reduce costs, speed up development, or improve success rates will have a significant impact on the financial status of pharmaceutical companies.
Hassabis pointed out that machine learning models can improve the drug discovery process in multiple ways. He believes that the enormous potential for saving time and costs is promising. However, he also cautioned that challenges remain in obtaining high-quality training data due to factors such as privacy regulations, data-sharing policies, and the costs of data acquisition. Nevertheless, he thinks these challenges are not insurmountable. Collaborating with clinical research organizations or using synthetic data can help fill the gaps in public data.
However, Hassabis emphasized that the application of AI in scientific research does not mean that scientists will be replaced. He pointed out that true innovation is still something that AI cannot achieve, as it cannot propose new hypotheses or theories. While AI can solve complex mathematical conjectures, it fundamentally relies on the wisdom and creativity of human scientists.
Additionally, companies like Nvidia are actively exploring the use of AI in drug discovery. Nvidia has even open-sourced the BioNeMo machine learning framework for drug development and partnered with several pharmaceutical companies to accelerate research progress.
Key Points:
💊 AI-designed drugs are expected to enter clinical trials by 2025, showcasing the immense potential of AI in drug development.
💰 The pharmaceutical industry faces challenges of high costs and low success rates, and AI is expected to significantly mitigate these issues.
🔬 Hassabis stated that AI cannot replace the creative thinking of scientists, and true scientific discoveries still rely on humans.
