- Capabilities
- Expertise
- Insights
- About
- Contact Us
en
Login
Update introduction to: We brought together CEOs from across the healthcare, diagnostics, pharmaceutical, life sciences, and biotechnology industries, to hear how they were actively piloting, adopting, and implementing AI technology. Here is what we learned.
Three key takeaways from our conversation with healthcare CEOs who are actively piloting, adopting, and implementing AI technology were:
|
The greatest risk of AI is inaction |
It’s important to move to a ‘factory approach’ to AI |
The learning capacity of senior leaders is key to AI success |
The risk is not leaping on the opportunity of AI quickly enough.”
CEO attendee
RRA x HBS AI CEO Labs
How healthcare CEOs are adopting AI
AI has the ability to narrow healthcare’s historical 20-year lag between innovation and the standard of patient care including the use of hyper-personalization and leveraging greater productivity across the ecosystem from R&D, to manufacturing and supply chain, to marketing. One CEO highlighted: “A key area of interest will be observing the integration of AI with the humanization of healthcare.”
CEOs also shared how AI is allowing the company to scale product development, and helping the company to achieve high levels of productivity and cost points. Another shared how Gen-AI helps drive Marketing productivity, allowing their company to advertise at 1/1000th of the historical cost.
Where AI thrives—and stumbles
In 2023, Karim Lakhani, a professor at Harvard Business School, launched a joint research study with Boston Consulting Group that looked at whether AI could tangibly improve the productivity and efficiency of elite knowledge workers. The study of 758 BCG consultants found that:
- Consultants who used AI were 25% faster and completed 12% more tasks than those who weren’t using AI tools. The impact wasn’t just felt from a speed perspective—the study also found that 40% of the group using AI also yielded higher-quality results.
- Consultants who were deemed below-average performers benefited most from AI augmentation—with an increase of 43% in performance. However, above-average performers also noticed a 17% boost in performance by using AI.
- There was a fall in performance when it came to using AI to complete complex tasks, such as the triangulation of financial data in a spreadsheet that contained errors or analyzing a retail strategy using interview notes. In these scenarios, consultants were 20 percentage points less likely to produce correct solutions compared to those who were not using AI.
Key lessons from healthcare CEOs actively adopting AI
Lesson 1: The greatest risk of AI is inaction
The change that AI will bring to healthcare organizations is vast, and entire business models will need to be re-imagined. Professor Tsedal Neeley at Harvard Business School shared how companies need to think about AI in terms of radical change, rather than incremental change. “It’s not about tweaking existing systems; it’s about creating new systems, processes, and structures,” she said. “In five years from now, our organizations will look very different.”
CEOs discussed that adoption of AI in healthcare is challenging, with risks around disintermediation, cyber security, the regulatory environment, the acquisition and retention of technology talent, lack of digital maturity, and culture change and resistance. One CEO also referenced the challenges of keeping the value in their company, rather than letting it leak to technologically savvy customers.
The main risk flagged was the pace required to keep up with the AI disruption. As one CEO shared: “This wave is moving very, very fast—faster than the mobile wave and faster than the Internet wave.” Attendees spoke about the threats of being radically disrupted, outperformed, or falling behind the competition, noting that the AI opportunity available to them was also available to their competitors. One CEO stated: “The risk is not leaping on the opportunity of AI quickly enough.”
Lesson 2: It’s important to move to a ‘factory approach’ to AI
CEOs highlighted how successful adoption of AI requires simultaneous changes in technology architecture and innovation processes. AI provides the opportunity to process incredible amounts of data and couple proprietary data with externally available sources (data twinning) e.g. scientific and clinical data, or multiple sets of patient records, which previously would have been analysed in isolation. It was accepted that traditional siloed data stacks must evolve into more integrated and agile frameworks to fully harness AI's potential—although this is only part of the equation.
One CEO shared: “We have gotten over the challenge of ensuring our data is organized in the right way, but having people that understand core business processes and what we're trying to accomplish with AI is a challenge. It's gotten so siloed for so long that we don’t have people with the broad perspective that’s needed to manage across historical silos.”
Two CEOs spoke of their radical tactic of building a new tech stack from scratch adjacent to the organization and integrating it, rather than opting for a protracted systems update and multi-year transformation. Building a brand-new team allowed both organizations to set up a whole new platform in record time.
One shared: “We started our AI journey about seven months ago. Bringing people along was very slow and resistance was high. Instead of trying to lean our processes, change our people, and build the tech stack in parallel, we hired one hundred new people. We now have the AI tech stack and the processes all in one place and we get to see how the future works. We’ve built this for three functions. If the model works, we’ll run it through a dozen other functions.”
Professor Lakhani emphasized that this kind of radical thinking takes guts. “We see many CEOs get caught in the sum-cost fallacy where they believe they’ve spent substantial time building systems and thus must be able to squeeze more out of them,” he said. “Instead, the mindset should be: at some cost, let’s go and build cheaper, faster, better.” While talent is scarce, with the pace of learning in some geographies we expect to see a rapid increase in available developers soon.
The AI Factory
An excerpt from Competing in the Age of AI: Strategy and Leadership When Algorithms and Networks Run the World by Marco Iansiti and Karim R. Lakhani.
The AI factory is the scalable decision engine that powers the digital operating model of the twenty-first-century firm. Managerial decisions are increasingly embedded in software, which digitizes many processes that have traditionally been carried out by employees. No human auctioneer gets involved in the millions of daily search-ad auctions at Google or Baidu. Dispatchers do not decide which car is chosen on DiDi, Grab, Lyft, or Uber. Sports retailers do not set daily prices on golf apparel at Amazon. Bankers do not approve every loan at Ant Financial.
Instead, these processes are digitized and enabled by an AI factory that treats decision making as an industrial process.
Analytics systematically convert internal and external data into predictions, insights, and choices, which in turn guide or even automate a variety of operational actions. This is what enables the superior scale, scope, and learning capacity of the digital firm.
Lesson 3: The learning capacity of senior leaders is key to AI success
AI is firmly seen as a CEO mandate, with RRA’s AI Practice lead Fawad Bajwa explaining, “Unlike the role of digital transformation, where you may have had someone you delegated this to, AI tech transformation at the level we're talking about lands squarely on the CEOs laps.”
CEOs are in the process of not only personally upskilling around AI, but also ensuring their senior leadership team understands its potential for business applications. Getting senior leaders more fluent in AI would allow them to think outside the box about how AI could be applied.
Professor Lakhani shared, “As you think about your own enablement, you will need a team around you that gets this in order for you to better connect with both internal and external stakeholders and run experiments. There has to be a basic knowledge not just about AI, but how it applies to business. It’s not about technology for technology's sake. It’s about how can we use it to create more value, or capture new value.”
CEOs agreed, with one sharing: “AI has to be a top-down mandate [from the CEO], driven by the functional leaders so that it flows through every part of the organization. The real question is how to get all the leaders in the organization to think about how AI can fundamentally change every part of the business and come up with their own use cases. We can’t be limited by what's out there today because it’s still early days. We need to find ways to get them to want to learn more about AI and to push the envelope.”
The 30% Rule
An excerpt from The Digital Mindset: What It Really Takes to Thrive in the Age of Data, Algorithms, and AI by Paul Leonardi and Tsedal Neeley.
A digital mindset is the set of approaches we use to make sense of, and make use of, data and technology. This set of attitudes and behaviors enable people and organizations to see new possibilities and chart a path for the future.
And here’s the good news: you only need about 30 percent fluency in a handful of technical topics to develop your digital mindset. We call this the 30 percent rule.
To understand the 30 percent rule, think about learning a foreign language. To demonstrate mastery of the English language, a nonnative speaker must acquire roughly 12,000 vocabulary words. But to be able to communicate and interact effectively with other people in the workplace, all they need is about 3,500 to 4,000 words—about 30 percent of what it takes to achieve mastery.
In practical terms, a nonnative speaker does not need to master the English language to work effectively with others. Similarly, to work effectively with a digital mindset, you don’t need to master coding or become a data scientist. But you do need to understand what computer programmers and data scientists do, and to have proficient understanding of how machine learning works, how to make use of A/B tests, how to interpret statistical models, and how to get an AI-based chatbot to do what you need it to do.