Powered by
English
Español
Français
English
Español
Français
Editor's note: This article was originally published in The Boardroom Magazine, November/December 2024.
On September 12, 2024, Open AI, the developer of large language model (LLM) ChatGPT and its various iterations, released its newest version, GPT-o1. What is different about this release is that a new artificial intelligence (AI) language model, Strawberry, was used. This model should improve the reasoning and problem-solving capabilities of existing language, which is currently in the generative AI that most of us have begun to make mainstream.
That said, while current generative AI models may not have a reasoning component, they do facilitate many tasks. One is the analysis of large amounts of data, as they provide a general overview of what the data is saying from varying perspectives in an easy-to-understand format.
Generative AI is a language model used to create bespoke content based on analyzing patterns, previous behaviors, and examples. Machine learning (ML) is one term used to explain the process. The technology learns or sees repeated behavior and mimics these when prompted. Examples of these are ChatGPT, Gemini, Pilot, and Poe.
The main stakeholders for clubs are their members, and ensuring they receive the level of service they expect is a top priority. Using generative AI offered in existing applications or as a stand-alone option can provide a valuable resource for management and staff to learn about the member’s preferences, past experiences, and usage or purchasing trends, allowing club management to be more proactive in anticipating members’ needs and creating personalized experiences.
How generative AI can be applied in a club environment varies. Some of the ways it is being used today are:
Communication: Many options in software exist now that help facilitate the development of newsletters, invitations, announcements, and club marketing. Such applications can allow clubs to speed up communication with members at lower costs without compromising quality.
Golf Course Maintenance: The use of drones to oversee areas of the golf course is not a new practice; however, now incorporating other generative AI applications, the data from the drone flights can be used to manage irrigation and maintenance requirements through systems communication, increasing efficiency and helping to manage costs.
Reservation Systems: Systems that incorporate generative AI can facilitate bookings of tee times, catering, restaurant reservations, or special club events. These systems remember member preferences and can cue members to quickly make their arrangements while providing operations with the information they need to plan the members’ experiences.
Prompting should be clear and specific. The process of asking for information is called ‘prompting.’ When using generative AI, phrasing questions or requested information needs to be clearly defined and specific to obtain the best outputs or responses. All users should learn how to write effective prompts.
There is a possibility of misinformation. ‘Hallucinating’ refers to responses received from generative AI applications that are not accurate, misleading or entirely made up. Often, the responses are presented clearly and appear factual, but upon further research, it is clear they are false. The number of hallucinations seen in an application can create distrust in results. Still, these can be diminished as more data is available on the topic at hand within the generative AI application or database.
Understand the risks of information-sharing. LLMs always remember, so be careful of the information you share within your prompts. For example, a club controller used Chat GPT to assist with their profit and loss (P&L) statement. The controller placed their entire financials into the system and was impressed with the analysis provided. Now the club’s entire financial picture (including the back-and-forth discussion the controller had about the analysis) is forever in the LLM’s memory – and it is also available to any other users who request this information with the right prompt.
Essentially, a modern LLM can function operationally as an unpaid intern if its capabilities are leveraged effectively. You can use it in many ways – perhaps for brainstorming ideas, assistance with Excel or writing emails, or creating staff schedules. But don’t become too overly reliant on it and always double-check its work. And if you look at it as an unpaid intern, think of it also as one who likes to gossip – so don’t offer it any sensitive information that you wouldn’t want shared with the public.
Given the impact of AI and emerging LLMs on the hospitality industry, HFTP will be providing multiple opportunities for club professionals to learn how to innovate with AI. It was the topic of discussion (“Innovate with AI: How to Leverage Large Language Models for Club Success”) at the HFTP 2024 Club Forum that took place October 22. HFTP has also developed a new micro-certificate to continue and further review the material presented at the Club Forum that is available to all interested club professionals. Learn more by visiting academy.hftp.org.