For our final panel discussion - "The Trough of AI Disillusionment: What have we learned?" - we were joined by experts from various sectors of scholarly publishing to share their experiences, challenges, and insights in implementing AI technologies. The session, moderated by Emilie Delquié (Chief Product & Customer Success Officer at Silverchair), included Kate Eisenberg (Senior Medical Director, DynaMed Decisions, EBSCO), Dylan DiGioia (Director of Engineering, Hum), Mohamed Elshenawy (CTO & Co-Founder, Sinai.ai), and Jeremy Little (Technical Lead, AI Transformation Team, Silverchair).
AI and tools like ChatGPT have fundamentally changed the technology landscape. Two years after the launch of ChatGPT we are now entering what Gartner's Hype Cycle refers to as 'the trough of disillusionment,' with interest beginning to wane as time passes. So what have we learned from early ventures into AI, and what are the early adopters saying to help inform our path forward?
Learning From Early Adopters
A poll kicked off the session, asking where attendees were on the AI journey. Results revealed that most attendees were in the experimentation and prototyping phase of their AI journey, with this option getting 40% of the votes. This was followed closely by those considering how best to use AI. This set the stage for a rich dialogue on the practical realities of AI implementation in scholarly publishing.One key point emphasized by speakers on the panel throughout this session was the importance of a measured, strategic approach to AI adoption. Kate Eisenberg emphasized the value of starting with a solid foundation:
"Being really disciplined about understanding where you are I think is important and not getting ahead of ourselves. Maybe you have a roadmap in mind, but understanding that you really do need to understand – does it make sense to take the next step just because the technology's been evolving so quickly?"The other panelists echoed this view and stressed the need for careful planning and realistic expectations when embarking on AI projects.
Challenges & Lessons Learned
The panelists shared candid insights about challenges they've faced. Jeremy Little from Silverchair highlighted the difficulty in navigating the rapidly evolving AI landscape and emphasized the importance of using due diligence and critical thinking when making decisions about new AI technologies and partnerships:"I think when we see new paradigm shifts and we see groundbreaking technology like LLMs and the open sourcing of many models, I think there tends to be a number of startups, new products, and companies that just blow up out of nowhere and get super high valuations and gain all this hype. And it's really difficult for us as a technology provider to sift through all of those new innovations and all those new companies without falling into the honey trap that is this amazing new feature that maybe actually isn't really all the way there."Mohamed Elshenawy pointed out another critical challenge: data quality and privacy. He noted,
"We found that many people who are using the system are not doing so the way we expected. And that results in a different output than we expect from the agent itself. We also had to deal a lot with the privacy issues and sign different agreements just to be able to use the data for experimentation."Both showcased the need for robust data strategies and careful consideration of privacy implications when implementing AI solutions.
Strategies for Success
Despite having some challenges with AI implementation, the panelists offered valuable insights on how the AI landscape can be navigated successfully. Dylan DiGioia argued that a structured approach is vital:"At Hum, we think a lot about the maturity and where we are at with every project or product. So are we ideating? Are we just thinking about what could be and the possibilities that we're going after, how can we apply some rigor to that, and is this product really going to work for us?"This methodical approach, moving from ideation through experimentation and evaluation before full-scale implementation, was presented as a key strategy for avoiding the "trough of disillusionment" often associated with new technologies.
Kate Eisenberg shared EBSCO's success with their cross-functional team approach, particularly in integrating clinical expertise into their AI development process. The strategy enabled them to remain agile and responsive to user needs as they brought their AI product to market.
Looking Ahead
How can organizations in the industry best explore AI applications? The panelists provided some valuable insights, including Jeremy Little, who cautioned against rushing into AI investments without careful consideration:"I think it's really important to stay out of this trough of disillusionment, to recognize what the role of your company is. I think that a lot of companies have been rushing into AI and throwing money and people at this problem when it doesn't necessarily make sense for that organization to be doing it."Instead, the consensus among panelists was to start small, focus on specific use cases, and learn through doing. As Kate Eisenberg put it:
"Start small, but do start somewhere. Because you can get that learning. And what we found is there really is a certain amount of learning by doing."The discussion concluded with recommendations for staying informed about AI developments, from following industry podcasts and conferences to curating professional social media feeds.
The insights these experts shared provide valuable guidance to other industry organizations looking to navigate the ever-growing landscape of AI. By approaching AI adoption with careful planning, cross-functional collaboration, and a willingness to learn and adapt, organizations can navigate the challenges and harness the potential of AI to drive innovation and value in scholarly communication.
Looking to discover more insights from Platform Strategies 2024? Explore the archive and experience the full recording of this session.