What AI Will and Won't Do for a Science-Based Business

Prepare to make the most of the opportunities AI enables by keeping your leadership in the equation.

Originally published on LinkedIn, July 21, 2023 (updated)

I wish I were twenty years younger to have a greater opportunity to use Artificial Intelligence (AI) in science. It will be able to assimilate complex data and enable discovery in ways we haven’t yet seen. It will help us gather facts and figures, analyze the medical need, and perform competitive intelligence that used to take weeks or months in days. However, this article isn’t about what AI will do for us in complex scientific endeavors; instead, it is to remind us that our leadership will still be needed and, in my view, more than ever. That is because we’re human and must deal with other humans to accomplish our goals. AI may sound empathetic in its responses, offering seemingly human-like pointers and advice. However, it won’t connect with our limbic system. That takes human-to-human interaction.

Daniel Goleman and colleagues contend that emotional intelligence is the most crucial aspect of leadership; without it, our leadership practices will be ineffective. Our limbic system is the seat of emotional intelligence. Significantly, it is something we are wired to respond to. In other words, our brains can’t help but respond. When someone is in a bad mood, that mood will spread to others. Conversely, if we find joy in our job, it is more likely that our team will experience it as well.

John Assaraf, a well-known neural marketing guru, teaches that people respond emotionally well before the rational part of their brain kicks in. If we don’t first engage how a person feels, they just won’t be interested in what we say, no matter how valid our argument is. We can immediately see where this principle could impact our ability to attract investors, but it goes deeper. For instance, we are more likely to consider results that pose a competitive threat or, conversely, substantiate our beliefs and pass over findings that do neither–potentially missing important information. AI will help bring that missed information to light, which will help. I’ve written about the power of engaging internal devil’s advocates to keep us open to new ideas, and I believe AI can be one of those devil’s advocates, but we will still have to watch the kind of queries we use. For example, we’ll still have to be open to challenging our beliefs as we judge the validity of ChatGPT’s results.

The impact of our leadership doesn’t end with engaging the limbic system. Intellect without comprehensive leadership skills will have a much more difficult time succeeding. Leadership skills will also help our science and the execution of our programs in the quality of the decisions we make, the timeliness of our actions, and the ability to stay true to purpose, for instance. Those that study leadership, like John Kotter (management of risk), Frances Frei and Anne Morriss (team motivation), Michael Porter (formulation of strategy), or Gary Pisano, who has studied science-based business, emphasize how various leadership skills impact results.

In science-based business, we are trained to rely on intellect and logic using our cerebral cortex, and we put a premium on it in both us and the people we hire. In medical translation, we are so focused on the scientific/medical proof that we are more likely than leaders in other fields to discount the need for leadership that addresses the human side of success. The rub is that no matter how correct the hypothesis is, poor execution will still cause us to fail.

Kotter teaches that the success of reaching our objectives is directly proportional to our leadership skills. In science-based businesses, the development risk is already high; upward of 75% or more pipeline programs will fail. While some of the analytical risks might be removed with the help of AI, the choices made and how a strategy is executed by people (not robots) will still be a deciding factor.

Top-performing staff will be motivated by, as Collins and Porras put it “Big Hairy Audacious Goals” that may use AI to determine the most promising goals and assess development risk associated with estimated outcomes (which I look forward to seeing in product and clinical trial design, for instance). Still, its assessment will not consider the mood, motivation, or skill of the people doing the work. I’ve always believed in the quote from Henry Ford “Whether you think you can, or think you can’t–you’re right.” Suppose people at the bench don’t believe in a program. In that case, we’ll have a problem with data: in the team’s ability to acquire it, have it be correct, and be of sufficient quality to reliably carry out a strategy, whether generated by us or AI. Belief in a program comes from the leader, which will engage both the limbic system and the cerebral cortex. And if there is a productivity slowdown within an R&D team, it will be up to the human leader to rekindle the passion and commitment.

So what does the integration of AI into R&D mean for our leadership? Perhaps more than ever, those who ignore their leadership’s impact will be overtaken by those who use AI wisely with empowered leadership skills. I predict that using AI without these skills will result in hardly more than any other technological advance. For instance, despite the increasing technology options in recent years, they have not substantially improved biotech’s success rate while adding to development costs, particularly without sound strategy and execution. While AI will help better use information to guide decisions and fuel discoveries, I’ve come to believe that it’s ultimately not about the information or the technology but how we choose to use it. That is where your leadership comes in.

Prepare to make the most of the opportunities by not taking yourself out of the equation.