Generative AI stands to add trillions of dollars of value to the world economy, according to a recent McKinsey report. The consultancy noted that about three-quarters of the value of generative AI spans four areas — customer operations, marketing and sales, software engineering and R&D. The latter could see a 12% annual lift in global functional spending as a result of generative AI, representing an approximately $328 billion windfall annually.

An evolution or revolution: UBS and McKinsey’s take on generative AI in biopharma

The investment bank UBS recently reached more muted conclusions, noting that generative AI would represent more of an evolution than a revolution for biopharma.

Conversely, McKinsey bets the technology could deliver productivity gains that account for 10% to 15% of overall R&D costs. This potential largely stems from the application of generative AI foundation models in what is known as generative design.

In the face of such perspectives, Adityo Prakash, CEO of Verseon, agrees with the transformative potential of AI but cautions against oversimplification. He underscores the need for a holistic approach to AI, taking into account the limitations of data. He points out a critical example in the conversion of chemical structure data. “If a lot of this data is just 2D chemical structures for a small molecule structure, what you’re supposed to do is turn that into 3D using tools that can generate a realistic 3D representation,” Prakash said. “The problem is that many people in this space often miss the details required for deploying AI.”

Beyond drug development: Generative AI’s impact on various industries

Indeed, many of McKinsey’s pronouncements extend beyond the pharma sector. The firm, for instance, projects the technology could automate tasks that now occupy about 60% to 70% of employees’ time across various industries. Ultimately, McKinsey estimates that “half of today’s work activities could be automated between 2030 and 2060,” roughly a decade earlier than in its earlier models suggested.

Prakash stressed that solving thorny drug development hurdles will require a cross-disciplinary approach. “To really solve the problem well, you need people who speak all of these languages fluently,” he said, referring to the various domains of expertise required in drug development – including data science, biology, chemistry, and AI. He suggests that a team combining a physicist, a mathematician and other backgrounds can potentially solve the problem better than siloed teams.

The aim is to expand what Lilly refers to as its “digital worker-equivalent workforce,” an idea intended to measure the time saved by employing technology in place of human labor. In an interview with Insider, Lilly CEO David Ricks stated that its initiatives, initiated in 2022 and encompassing more than 100 projects, are comparable to nearly 1.4 million hours of human effort, or roughly 160 years of round-the-clock work.

In response to the pharma giants’ embrace of AI, Prakash suggested that while these steps are commendable, it’s vital not to underestimate the challenges ahead or overstate the capabilities of AI alone. “We need to get there, and it’s a completely non-trivial task. AI is only one of many tools that are going to help,” Prakash said. “Just chanting ‘AI’ doesn’t get you there,” he stated.

---

Filed Under: Data science, Drug Discovery and Development, Industry 4.0, machine learning and AI
Tagged With: ai challenges, AI in Pharma, digital workforce, drug development, generative AI, mckinsey report, productivity gains, Verseon
 

---