How have you reacted to the buzz about generative artificial intelligence (AI), especially since the release of ChatGPT last November? If you are like some in the learning and development (L&D) community, you’re skeptical about the early chatter. On the other hand, you might be an early adopter or even a power user and enjoy sending daily prompts.
As part of the ATD Forum’s mission to connect, collaborate, and share, over 50 members recently gathered for a roundtable discussion on generative AI, hosted by member Dana Koch of Accenture in discussion with thought leader Elliott Masie. The two previously collaborated on a podcast. According to Techopedia, “Generative AI is a broad label that's used to describe any type of artificial intelligence (AI) that can be used to create new text, images, video, audio, code, or synthetic data.”
With 61 percent of attendees not using the technology, there were many comments like, “I don't know if I am amazed, or terrified, or both :).” However, with 39 percent of attendees being users, so many contributed additional resource ideas and lessons from their own experiences.
Masie and Koch provided many insights, helpful hints, and prognostications. Masie mentioned that while the buzz is new, AI is not. There were versions of a chatbot in 1971. However, currently, there is a convergence of the technology, a case study for use, huge investments of money by many tech companies, and inherent competition that’s pushing the market.
According to Masie and Koch, there are limitations with the current versions, especially related to tagging and ownership of the data. However, these kinks are being worked out with each new release or product. For L&D to be a driver, we must join with internal partners, especially IT, and remain aware of the capabilities, experiment with technologies, and learn from the experiences. Some current uses for generative AI include customer service, sales, health care, and coding.
Ideas for moving forward:
- Write effective prompts to brainstorm a new course.
- Educate others on AI’s uses to help them overcome fears of the technology.
- Learn about the potential power of the technology to deliver next-generation performance support.
- Embrace the shortcuts in brainstorming and writing; work alongside this tech.
- Think about how AI can help realize learning in the flow of work.
- Use AI as a launching point for when people feel "stuck" on next steps—craft a resume, develop plan ideas, etc. Then refine after your words are on the paper.
Considerations mentioned in the discussion:
- Prompts are used to determine the output—skills are needed to understand which prompts work for your needs.
- The trust factor matters, especially in how much people need to review generative AI outputs.
- There are legal implications around sourcing, tagging, copyright, and ownership.
- The technology goes beyond text-based interactions.
- The need for an employee-use policy is inherent, especially since the prompts are permanent.
Arvind Narayanan’s recent Robert Tucker blog quote is pertinent to this conversation: “The danger is that you can’t tell when it’s [information] wrong unless you already know the answer.” Can we in the L&D space provide responses and solutions using technology—but not have the knowledge or skills to share how we arrived at the solution?
Several attendees provided reflections related to the session:
“We’re now at the point where all the incremental AI research and engineering improvements produce outcomes that seem like magic. Like in the early days of the internet, the race is on to see who can make the fastest use of amazing new capabilities. Areas like customer service seem like an obviously limited fit, while the problems of accuracy, explanation, and ownership loom large. During this time, it is critical to connect to other professionals and organizations working to progress in this space. Every new experiment will spark ideas for experiments that could be run elsewhere. Groups like the ATD Forum provide a valuable space for this innovation-enabling sharing.”
“Don't rely too much on AI until it stabilizes and gives accurate, unbiased information (who determines what sources are used) and ethics (tagging and providing sources for the content). However, the future of AI is bright, as I see many companies investing in it.”
“Generative AI/ChatGPT has been at the outer edges of my awareness—knew it was there and had a general idea of what it did. However, seeing firsthand how quickly the technology is evolving was eye-opening. I was shocked, for example, at how rapidly and completely it could create course outlines based on the criteria provided.”
“The idea that it could have broad application beyond simple decision-tree interactions in, say, a customer service context, squares with my thinking and was definitely reinforced during the session. My concern, though, is that it will become (or won’t evolve beyond) a high-end version of Wikipedia and the often-faulty information resident therein.”
“My questions to ponder: Will it ever be within reach of organizations to maintain an internal capability linked exclusively to their internal sources of content, and, if so, when will that be possible given all the legal issues?”
“What effect will it have on L&D roles? I can imagine a future where applied work/learning research professionals largely supplant the traditional L&D role. One intriguing possibility associated with such a future state is that AI would become a nexus that integrates research-based learning designs, line-of-business expertise, and on-demand performance support. Unfortunately, that would likely relegate traditional learning-related approaches to the dustbin of history.”
As a vendor-free consortium, the ATD Forum is an excellent playing field to share resources like recent blog posts, podcasts, articles, platforms, apps, and experiences. It is a safe venue for sharing the good, the bad, the ugly, and the pluses and deltas of experiences. With a diverse group including early adopters, those tiptoeing into the space, and those unsure what to think, there are varied ideas and resources, especially around topics such as generative AI.
The big takeaway: talent professionals live in a world of many disruptions, and generative AI is just one of them. So beware of the consequences of your decisions and actions, and be aware of the vast possibilities!
For more information and resources, checkout ATD's AI resource page.