Recent advances and investments in artificial intelligence (AI) have been making major headlines lately. As MRP highlighted last week, Microsoft’s potential investment of $10 billion to expand their stake in OpenAI, the creator of popular AI tools ChatGPT and DALL-E, could more than double the startup’s valuation to $29 billion. OpenAI’s suite of products, now set to go premium for use by corporations, utilizes generative AI, which produces unique images, text, and other mediums of communication from human prompts. About 175 billion machine learning (ML) parameters make up the deep learning neural network used in OpenAI’s latest model, GPT-3. To put things in perspective, Analytics Insight notes that Microsoft’s Turing NLG model, which has 10 billion parameters, was the largest learned language model before GPT-3.
Data and parameters used in neural networks and their resultant outputs can be focused on different subjects and industries. The healthcare industry, particularly biopharma and diagnostics, is a business where generative AI is expected to create major disruption in the years ahead. Per survey data published in a recent GlobalData report, The State of the Biopharmaceutical Industry – 2023, 39% of surveyed healthcare industry professionals believed that AI would trend as the most disruptive emerging technology in the sector throughout this year. AI has been voted as the most disruptive emerging healthcare technology every year since 2020, according to Pharmaceutical Technology.
Just last week, BioNTech SE agreed to acquire British artificial intelligence (AI) startup InstaDeep for the equivalent of up to $682 million, the firm’s largest takeover deal ever. Per the Chemical Engineer, BioNTech’s acquisition will support the company’s strategy to build capabilities in AI-driven drug discovery and development of next-generation immunotherapies and vaccines.
Several months ago, Nvidia announced the launch of its BioNeMo Large Language Model (LLM) service to help researchers build new artificial intelligence (AI) models for biology, which was used to build a new generative AI model that could have a significant impact on drug discovery. Working alongside biotech startup Evozyne, Fierce Biotech notes the pair showed they could use their program to add dozens of amino acid mutations to a human metabolic protein known as PAH, changing its shape into a more efficient form and increase function of the proteins. AI accelerated engineering of proteins cold help to significantly improve the efficiency of new drug discovery.
In the realm of diagnostics, European drugmaker AstraZeneca has teamed up with Indian digital health specialist Qure.ai and a clinical group in the UK to test whether AI-powered technology can help radiologists detect lung cancer in chest X-rays more quickly and accurately. As PharmaPhorum writes, Qure.ai’s qXR software, which has been trained on more than 2 million X-rays and their corresponding radiology reports to detect and localize up to 29 abnormalities, will scan more than 250,000 images from a British hospital in an attempt to carve precious days or months off of the time it takes to accurately diagnose lung cancer in patients.
Pfizer launched a similar partnership with AI-focused Anumana in 2022, working to develop programs that can help diagnose cases of heart disease like cardiac amyloidosis, where misfolded proteins build up around the tissue. Anumana believes there is a wealth of data still to be mined from reading into the electrical activity of the heart.