Artificial intelligence (AI) has slowly been integrating itself into our daily lives for quite some time now. You’d be forgiven to think that AI is a rather recent phenomenon with the advent of Tesla cars but astonishingly, the earliest successful AI program was written in 1951 by Christopher Strachey (as referenced by the encyclopedia Britannica). The University of Manchester based Strachey created a checkers program that was shown to successfully play a complete game of checkers at a reasonable speed by 1952. Fast forward 70+ years and Elon Musk is spearheading a humanized Tesla robot while Jon Snow is making a passionate apology speech for season 8's pitifully dismal ending. Not only is AI grabbing headlines for its salacious use of deepfake, but it’s also making history. In March 2016 AlphaGo, a computer program utilizing deep neural networks, was the first to defeat Go world champion Lee Se-dol in a 4-1 victory; prompting Lee to recoil into retirement stating “Even if I become the number one, there is an entity that cannot be defeated”. Given AI’s recent advancements, it comes as no surprise that it has nestled itself into the biotech/pharma arena under the guise of “AI powered drug discovery” with promises to develop better medicines faster and cheaper.
Artificial intelligence, or rather, machine learning (ML), is the hottest thing to hit the biotech market since the Carl June inspired CAR therapy days of the 2010’s; everyone is trying to do it and everyone wants a piece of it. According to an article titled “9 Notable AI-powered Biotech Companies Founded in 2021”, AI driven biotech companies have seen more than a 7 fold increase between the years of 2010 to 2021, with the biggest surge in companies founded between 2016-2019. Additionally, AI biotechs saw a huge surge of VC investments, in both early and late stage in particular from 2020 onward as evidenced by this chart from McKinsey and Company
Amidst all the flashy buzzwords and excitement in this space, there remains some confusion over what exactly AI/ML is within drug discovery specifically, which companies are using it, at what point in the drug discovery process, and has it actually created the “magic bullet” or as Paul Elrich coined: “Zauberkugel”? As a friend in the space told me recently: “Truth be told, it’s very difficult to know which companies are actually using the technology by their glossy websites and popular buzzwords”.
In this three-part series, we will take a closer look at how AI/ML is being utilized within drug discovery, who are the big players in the field, and who are the big investors in the space. Before we delve into the details, let's first discuss what AI/ML is.
You’ll often hear the terms AI/ML used interchangeably in the context of drug discovery but there is indeed a difference between the two. Columbia’s engineering page explains it quite well: “Artificial intelligence refers to the general ability of computers to emulate human thought and perform tasks in real-world environments, while machine learning refers to the technologies and algorithms that enable systems to identify patterns, make decisions, and improve themselves through experience and data.” Machine learning is a subcategory of AI, a pathway to AI if you will, and seems to be the most common subset used in biotech. Other notable subsets include deep learning, neural networks and natural language processing. The intentionality behind AI and its subsets is to organize, automate, and streamline manual processes involving large sets of data for faster and more efficient output so it’s no surprise that the most common places AI has been utilized are in the manufacturing and healthcare industries, and it is even less of a surprise that drug discovery wants in on it.
Next month we will examine the companies who are famously using AI/ML in drug discovery, at what stage they’re utilizing it, and has an AI powered drug been discovered? The answers might surprise you.
(Disclaimer: this blog was not written by ChatGPT but by an actual human. A quite nice one at that!)
All sources have been directly linked into the blog post.
