GenAI: How Disruptive, How Soon?
At an upcoming MIT event experts will offer leaders advice on letting go of control and optimizing AI
The impact of GenAI is still a developing story. Some say that GenAI has already moved the needle for the digital economy; that the next major innovation disruption is here now. Smart healthcare apps, for instance, could generate as much as $1.2 trillion in value by next year, according to some predictions. But there are concerns, as well.
Is your business prepared for this transformational change? Or are we still in the very early stages of development?
Several expert speakers at the upcoming BigAI@MIT conference to be held at MIT November 15 will examine these issues so that business leaders can get on board more fearlessly. IDE Editorial Content Director, Paula Klein, asked Des Dearlove, Co-founder at Thinkers50 — a co-host of the event along with the MIT IDE and Accenture — and two of his panelists, for a sneak peek of their discussion, GenAI: Disruptive Innovation or Paradigm Change?
Based on conversations with Thinkers50 leaders, Dearlove sees businesses facing huge changes that involve balancing machines and people, workforce transformation, and ethics. At the event, “we want to put AI in a historical context and help businesses make sense of it all,” he said.
Rita McGrath, is an author, strategic management scholar, and professor of management at the Columbia Business School. AI gains are impressive, but she cautions that you can’t really talk about AI in isolation, you have to get specific. McGrath specializes in strategic inflection points and navigating business change. Her new book on “permissionless organizations,” studies how corporate hierarchies will change as decision-making is pushed to the edges tapping into underutilized resources.
Another panelist is Scott Anthony, author, innovation advisor, and professor at the Tuck School of Business at Dartmouth College. Anthony has deep expertise in disruptive innovation theory based on his work with the late thought leader, andThinkers50 Hall of Fame member Clayton Christensen. Anthony’s upcoming book, tentatively titled Anomalies Wanted: How Disruptive Innovators Change the World, will be published by Harvard Business Review Press. “There is a heck of a lot of behavior change that has to happen to have AI adopted at scale,” he said.
What follows are perspectives that will be highlighted during the panel:
Des Dearlove: We hope this panel will add context to the discussion of GenAI. We want to address whether it is a game-changer or a new game. We’ll also explore how new organizational ideas and practices may be needed to meet the radical technology challenges ahead.
Rita McGrath: First, I should note that talking about AI in isolation is not all that useful. It’s sort of like talking about “the Internet” or “electricity,” in isolation. The value is going to depend very much on specific use cases, and we are still in the early days of figuring out what those are.
That said, there are any number of ways in which AI can be applied to remove frictions and free up value — and I don’t think the biggest impacts are going to be in making white collar professionals a few percentage points more efficient. Already, we are seeing fascinating ways in which AI can be used. In an example of counterfactual reasoning, for instance, when California raised the price of a pack of cigarettes, researchers were able to model what would have happened without that policy change and quantify the health impacts. AI is also behind smart healthcare apps for medical diagnoses. Pundits are predicting that as soon as 2025 smart healthcare apps could generate as much as $1.2 trillion in value.
Additionally, AI smart agents can make lightning-fast decisions about things like insurability, credit-worthiness, and identity that would have taken humans longer and been less accurate.
On a broader scale, AI holds the promise of vastly reducing the frictions we all face in using digital technologies
— passwords, two- factor authentication, filling out complex forms multiple times to record unchanging basic information, and more. So I think we’re going to see a rewiring of much of the digital infrastructure we already use.
Despite all of this progress, the vast majority of large organizations are not yet ready to truly benefit from the potential of digital technologies in general, and AI in particular.
In the worst cases, think of a traditional bureaucracy where siloed business units won’t share information, tech is siloed in the IT department, and most of the applications are either massive enterprise-wide applications or thousands of local apps that don’t talk to one another.
By contrast, in the most sophisticated firms (Novartis, for instance), access to digital tools are democratized and people are trained to use them; technologists partner with business unit leaders, and the fundamental unit of activity is a small team where problems are tackled end-to-end in rapid cycles. This creates the potential for dramatically increasing speed and making the “permissionless” idea of solving problems at the front line a reality.
Additionally, leadership needs to emphasize enterprise-wide thinking, not local silos, and have the courage to leave behind practices and rules that no longer fit for purpose. Radical simplification of work may well be in order, as MIT’s Zeynep Ton and other experts have recommended.
Leaders will have to give up many traditional forms of control, but they have the potential to gain enormous amounts of transparency and effectiveness in exchange.
Scott Anthony: It is important to recognize that while it feels like things are happening at lightning speed, work on AI began about 70 years ago. Therefore, it’s not that surprising to say that 70 years in a technology is well beyond the pure exploratory stage!
I fully agree with [Rita McGrath’s] point that you can’t talk about AI as you would talk about the Internet. You have to get specific. Specific tasks, use cases, and contexts, are what really matter.
Some tasks have already seen transformational change (scripts for call center agents, for example), and other tasks are in earlier stages (i.e. drug discovery). And in other tasks it is not clear that current AI will ever be the “answer.”
Also, when we talk about disruptive impact, it’s important to be clear on vocabulary. I consciously use the word “transformational” change as a way to say “holy heck this is BIG.” But I reserve the term “disruptive” for the Clayton Christensen sense of a paradigm shift that changes the basis of competition in ways that make it challenging for incumbents to respond naturally.
This clarification is important, because in some contexts incumbents will be well positioned to manage AI-driven changes that enhance productivity and enable new offerings to consumers. In other contexts — think education and professional services — realizing the potential of disruptive AI-grounded approaches will really require very careful, thoughtful, active management.
Regarding AI adoption, I reflect on the lesson learned by Paul Cobban, the Chief Data and Transformation Officer at DBS during its digital transformation. “Nothing changes,” Paul would remind us, “unless people’s behavior changes.”
There is a heck of a lot of behavior change that has to happen to have AI adopted at scale.
Some of that change is scary because it involves people giving up power, which they don’t naturally like to do. Some of it is hard because it requires unlearning old things before you learn new things. Leaders that don’t pay attention to the human side of this change are going to find it much harder than they think.
One needed change in mindset may be to give people permission to play around with GenAI and other emerging technologies. Adults have sadly forgotten the benefit of free play, and rather than approaching it with glee, they approach it with fear. The more you inject fun into it, the more you can learn.
The MIT Initiative on the Digital Economy, Thinkers50, and Accenture are joining together to co-host the first BIG.AI@MIT event on November 15, 2024. Apply to Register Now.
