Conversations: Mike Bushman on Generative AI for Events — Trust, Transparency, and Notable New Names 

After steady buzz about the capabilities of ChatGPT and generative AI last year, we’ve since seen countless leaps forward — along with potential new reasons for caution. RainFocus CTO Mike Bushman has kept a steady eye on the space and expertly evaluated each development. Last year, he and other RainFocus executives weighed in on the hype vs. game-changing potential. 

We’re excited to bring Mike back for another wide-ranging, candid discussion of what the events industry should know about generative AI in 2024. We covered everything from the impact of famous new entrants to the most common current real-life use cases your peers may already be deploying in their event planning.

The generative AI space is fast-moving. Over the past year, what do you think have been the most significant advancements?

First, there are more players out there, and others have moved ahead in their releases. AWS Bedrock went from preview to generally available. Having another major cloud player in the mix helps a lot. There have always been many questions around IP, trust, and governance. So, being able to use a trusted provider is really valuable here. 

We’ve also seen advancements from other big names. There are new models from Google, and new capabilities from Oracle in their cloud. Of course, the models themselves have become more advanced — ChatGPT has had new versions, and Anthropic Claude 3.5 came out as a compelling model. 

Seeing highly recognizable names enter certainly has been huge. We recently saw another big company make a splash. What do you see as the effects of Apple’s recent announcement of its Apple Intelligence? 

Apple’s introduction of a smaller model directly on the device is significant. Having processing speed on the actual device without relying on cloud services will be a big advantage. Apple has also done well so far in telling users about what the device can and cannot do, and when questions will be answered off the device. That transparency is important. We’ll have to wait and see the full extent of the broader impact. 

Exactly. As we’ve discussed before, accuracy and trust continue to be concerns. A recent occurrence that got a lot of attention was Google’s AI-generated summaries of search results that wound up showing completely inaccurate information. In some cases, “answers” pulled from satirical news publications were dangerous if interpreted as fact. 

For companies, what are some best practices for both content creation and guarding their own reputation when it comes to content out there about them?

One initial thing many companies are doing is putting clear liability statements out there. Covering yourself on that level is key.

User feedback will also continue to be very important. A simple and effective UX makes it more powerful. We’ve seen clear “thumbs-up/thumbs-down” options and the ability to validate content through personal verification. That human interaction to show that content is trustworthy is key. 

On the product side, of course you need to have proper tests in place. Sentiment analysis and others should always return solid results. We’ve even seen times when the generative AI output quality is suddenly wildly different even when a new model has not been released. 

In any case, research is critical. Often these answers found through AI search summaries provide new insights you were not aware of. They might not be taken at face value, but if you’re including them to seek external validation, it can be very useful.

Let’s talk more about generative AI use cases in events and where we stand. In our events and marketing survey currently underway, early respondent results have interestingly shown that so far, the majority of people are not using these tools for their events yet. What might be some reasons for the hesitation? 

Last year, the main focus was developing proofs of concept and verifying technical viability. Understanding costs was another big part of it. As major cloud providers like AWS were still working on their early releases, many people were still in wait-and-see mode. 

This year, we do see more companies in events actually implementing it — or at least putting it on the roadmap. Next year will bring in many more companies jumping in. Among event tech providers, it’s a similar story. Some are dipping their toes in. More will definitely embrace it over time. 

Stay tuned for Part II of this fascinating discussion with Mike! We’ll delve into additional use cases for events and get a sneak peek at RainFocus’ own AI innovations in progress within the platform.