AI Is Coming
May 9, 2019
By Dan Wacksman, CHDM, Principal, Sassato, and chair of HSMAI’s Marketing Advisory Board - Original Article published by HSMAI
Artificial intelligence (AI) is rapidly evolving, and even in these early days it is already having a significant impact on business — not only through its application but due to the attention now being paid to it, from board rooms to brainstorm sessions. As AI becomes an increasingly utilized technology, it becomes more important for us to understand the potential implications it will have on our business, both directly and indirectly. During a recent call for HSMAI’s Marketing Advisory Board (MAB), I led a conversation about AI, based on a presentation I gave during HSMAI’s Spring Curate 2019 event in Nashville. Here are a few takeaways from that discussion:
1. We have to be able to understand AI to utilize it. “I do feel passionately about the fact that AI is dramatically changing the way that we do things, and that we can either wait until other people figure out how they’re going to take our data and charge us for it, or we can educate ourselves and get into it and spot some opportunities on our own,” one MAB member said.
The group agreed that AI can be difficult to define because often people have different senses of what it is, including what technologies qualify. One MAB member defined AI as “a machine that learns and can improve what it’s doing.” Words like machine learning, deep learning, and neural networks seem to confuse people, but the MAB reviewed the chart below and agreed that this is probably a good illustration to help people understand the different components of AI. But still, what actually qualifies as AI is often a matter of contention. Is it a simple decision tree or artificial intelligence? At the end of the day, whether something is defined as AI or not, the most important thing to consider is what business problem are you trying to solve and what is the right tool to utilize to achieve this.
2. There will be challenges. AI is the current buzzword in hospitality and beyond, but the reality is that we’re often not doing the basic blocking and tackling that will enable us to get to the effective use of AI, which will only be as good as the data that it is fed. A big problem in the hospitality space is a lack of clean data, because a lot of hotels are working on legacy systems with other systems bolted on top of them, often not talking to each other. While artificial intelligence is the goal, we are at risk of creating “artificial stupidity” with bad data and wrong assumptions. We must get the foundation set before we can go to the next level. Another challenge for the widespread use of AI in our industry (and others) is the lack of qualified data scientists, which will become an increasingly valuable skill set. There was debate among the MAB as to whether hotels will be able to develop in this space or will need to rely on third parties.
The success of AI will be incremental, and while people are excited about the sexy robots predicted by science-fiction writers, the next several years will be more focused on narrow AI, which is AI utilized for very specific and narrow functions. As Bill Gates once said: “We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next 10.” It is easy to laugh at some of the attempts at strong AI, like the robot hotel in Japan, but it will be the combination of the incremental advancements of narrow AI and the dreamers who push the envelope that will lead to that “rEvolutionary” change.
3. There are infinite possibilities. “Every new management system has some level of machine learning in them if they’re using forecasting that is learned from events and pieces of data that don’t make sense,” an MAB member said. “The guest side is really interesting, because we are getting much more comfortable with using AI, like Alexa, in our everyday life. So, what pressure does that put on our hospitality environment? How do we respond to that and yet still deliver on good customer service?”
One AI application that many hotel companies offer — which depending on the implementation may or may not truly contain AI — is chatbots, which allow customers to ask basic questions about products and services, with the bot supplying answers. Other companies like Expedia are attempting to develop more complex applications that allow customers to find and book travel by talking to a bot, just like you would a human. Taking that even further, Google Duplex will call businesses such as restaurants or hair salons and make a reservation for you. Testing these applications out, you will note that they don’t work as well as the PR videos would have you believe, but it is easy to see that over time these will improve and spawn other applications that will have a huge impact.
MAB members agreed that a key area where AI could be used in the hospitality industry is for revenue management. If system can take in key data points from past history, weather patterns, flight patterns, events, compset, and so on, you could potentially come up with an automated pricing system. From a marketing perspective, attribution modeling also seems to be an area that could be positively impacted by AI, creating models that actually show and/or predict the value of each ad placement.
The MAB agreed that over time AI will impact every aspect of our business (and our lives) and will be important for us all to follow, learn, and, when appropriate, implement.