AI in travel and hospitality and the ethics behind it. We speak with Emma Taylor, Thematic Analyst, and Nick Wyatt, Head of Research, from GlobalData as we delve into the impact of artificial intelligence, how to be prepared and the dangers of implementing it without the right considerations.
1. How AI is being deployed effectively in travel and hospitality
2. The quick wins for companies exploring AI and the benefits delivered
3. ChatGPT and the use cases for generative AI in the industry
4. Ethical frameworks and sustainable data practices
Emma and Nick spoke at the International Hotel Technology Forum in March 2023 in Vienna.
A letter co-signed by Elon Musk and thousands of others demanding a pause in artificial intelligence research has created a firestorm - calling for a six-month hiatus on the development of systems. The focus is on the deployment of large language models. These models, which can predict the next word in a sequence given the preceding words, are the backbone of many conversational AI systems. However, the experts argue that these models are more powerful than ever before, and they simply don't understand how they work. This has led to concerns that the models are creating their own internal goals and behaviour patterns, which (they) are unable to understand or control and may exceed human capabilities
Listen on Spotify
Listen on Apple Podcasts
Programme Notes
Ryan Haynes:
Hello and welcome to Travel Market Life. I'm your host, Ryan Hayes. And today's episode we're going to be looking at AI in travel and hospitality and the ethics behind it.
Ryan Haynes:
Joining me now is Emma Taylor, thematic analyst and Emma Taylor, head of research at Global Data. Hi, thanks ever so much for joining me today. Now, I mean, AI is a massive topic in the industry right now. I mean, you can't turn to a magazine, a newsletter, or social media, without actually seeing it in a stream of some kind. So, it is a tech of the moment. I'd love to know from you guys, as you're looking at the industry and seeing how they're sort of engaging with this technology, in what ways is it being deployed effectively within travel and hospitality? I mean, Nick, I know you've been particularly exploring this.
Nick Wyatt:
Yeah. Well, first of all, Ryan, thanks for inviting us on and yeah, you're right. It is the technology of the moment. But I think, just to take a step back, I think we have to look at what we actually mean by ai because it's quite a catchall term. You know, we could be looking at things like large language models, we could be looking at other forms of machine learning, we could be looking at computer vision, we could be looking at conversational platforms. So, there's a lot going on and I think that there are examples of where, you know, it is being deployed effectively. And I think the most basic example that we see is where maybe certain repetitive tasks are being sort of automated. And in terms of hospitality, the advantage of that is that you're freeing the sort of staff up to spend time with guests.
Nick Wyatt:
As we all know, the kind of, the number one thing you're striving for is a high level of guest satisfaction. That guest experience is at the heart of everything that hotels are doing. That becomes arguably more important the further up the sort of value chain you go within, within a sort of star rating system if you like. So, I think that that is one area we see certain sorts of tasks, maybe around check-in, for example, is one, managing bookings, those kinds of things. Using AI in those regards is something we're seeing more of. I think the other element is personalization. Being able to make recommendations either for upsell opportunities on an existing stay or perhaps for, you know, post-stay promotions, you know, individualized promotions after that say, okay, we know that from the data we've collected, analysed, et cetera, this person stays in this type of room with this view.
Nick Wyatt:
They've had a tendency to, for these dining options. You know, can we now market accordingly? That's something that we've seen. But I think what, in terms of sort of the hotel industry specifically, there are I think a couple of industries they could learn from. I think we've seen cruise industries do some really interesting things around sort of ai. If you think about something like Ocean Medallion on Princess Cruises where it's used as, you know, the way somebody pays it's used as their dorky, it can, there's a whole lot of sort of information it can gather on the guest, and I think that's something that hotels are looking to move toward. And the other one I would say is I think there are some really good uses of AI with regard to dynamic pricing in the airline industry.
Nick Wyatt:
I think there's a lot that goes on with sort of pricing IQ sort of models within aviation, and I think aviation kind of has sort of set the standard where that sort of form of AI is concerned.
Ryan Haynes:
I mean, we always sort of look at technology and deployment across the travel and hospitality space and as you say, where we sort of see the real leadership and innovation is within aviation and how they're sort of drawing on how it's being deployed in other areas like entertainment and as, as, and retail. And obviously picking up that point there around dynamic pricing, Nick, it's fascinating and the power it can offer actually from a personalization per perspective and really sort of address some of those points of loyalty and really drive that consumer demand. How far do you feel that the hospitality is actually at reaching that stage?
Ryan Haynes:
Because I know when we are at the International Hotel Technology Forum, you know, there is always, there's a lot that conversation of, well, the industry isn't just there yet. There are a lot of issues around even just making sure the tech stack is in place and that there are those integrations. Nick, just AI rely heavily on the need to have these systems integrated and data to be shared.
Nick Wyatt:
It does, to do it effectively. I think that's the key. I think, you know, collecting data is one thing, then actually doing something meaningful with it and realizing that some of the data don’t really offer you a whole lot. You know, I think somebody, one of the speakers used the analogy, didn't they, that it, it's like water in the sense that, you know what, 70% of the Earth's surface is covered by water. A very small percentage of that is actually fit for consumption. And they said that data is very similar to that. So yes, I think they are. I do think that the hospitality industry has made good strides in this area and I think that it is interesting you mentioned loyalty because I think that that's been a driver behind it because obviously, we've seen hotels want to move more towards, you know, direct bookings so that they're not having to be so reliant on OTAs.
Nick Wyatt:
Obviously, there are commissions involved in that, et cetera. Then again, there are then some failings. I mean, you know, I had a conversation with somebody at the conference who was just still completely dumbfounded by the lack of price parity between these platforms. The reason that a lot of people put from OTAs in this person's view was they often see it cheaper on there, you know, and they're not going to, it's all well and good a hotel chain saying, well, if you get in touch with us, we'll match it. People are not going to go out of their way to get in touch with you. It has to be there at the point of exploration, at the point of planning. You know if it's not there then you've probably lost them. So, I do think that it is moving in the right direction, and I think loyalty is a good way of, of kind of doing it and trying to drive direct bookings.
Nick Wyatt:
And I think, you know, as with all things, some companies are doing that better than others.
Ryan Haynes:
And Emma, I mean you were given a presentation at IHTF around the ethical dilemmas of ai, but before we actually address that, what do you see within the industry? Are there quick wins for companies exploring artificial intelligence and, and the types of benefits that they can be delivered?
Emma Taylor:
Yeah, well I guess Nick, you know, obviously like just explained a lot of them, you know, automating a lot of that processes and really and really freeing up the time of, of, of workers in the industry and creating that better booking experience as well for customers fostering, increasing that loyalty. I think things like increasing accessibility to customer service, that kind of thing, and just immediately having a 24-hour chatbot that works well in terms of cybersecurity as well. Nick gave a good talk on great talk on cybersecurity at the conference and AI as well can be, can be used for, you know, fraud detection prevention. I mean really at every level I think there are small things that can be automated for operational, yeah, efficiency and cost-cutting even.
Ryan Haynes:
Yeah, so I mean I guess it's about looking at those solutions and identifying where they say the benefits are. But you know, the big conversation Emma, that we're having at the moment is around chat G B T now, what kind of use cases do you see for this generative ai, particularly within travel and hospitality? Because that's a big question on everybody's lips. I mean it creates articles, but how can we use it within the industry right now?
Emma Taylor:
Yeah, definitely there's such, such hype around it and such a sense of urgency for everyone to really start to deploy something. And I think there are a lot of really valid use cases that I could see for the travel and tourism industry. You know, as I said, with customer services, customer service in general. So, you could get ChatGPT or another generative AI large language model and integrate that into a hotel, your hotel's website or an app that then would then provide, you know, 24/7 customer service or could, could reply to frequently asked questions about the hotel and the hotel services and then what we mentioned earlier as well in terms of personalization.
Emma Taylor:
So ChatGPT could then be used to recommend room types, suggest nearby attractions, and that kind of thing for guests staying at a hotel. And then in terms of operations, so as I said, operational efficiency, I saw one article which explained how you could then integrate it to help with things like housekeeping schedules and inventory management. So then, you know, it could chat to keep GPT could sort of send you an alert to let you know when a room needed cleaning. And then, yeah, of course as you said mentioned, you know, it can write articles. So, in terms of marketing, I can see, see it working there.
Emma Taylor:
And then I think the final thing, which I think is probably one of the most interesting is in terms of product development. So I think ChatGPT and other models are now, they're now creating these generative AI app builders, which means that people with less technical proficiency can build different platforms than they could have before. So, you know, they could now create bots or chat apps or digital assistance using generative AI instead of needing Yeah, so much technical know-how,
Ryan Haynes:
I mean it is fascinating there just making everything so much simpler and easier and quicker to be able to move things and evolve things much faster. And I certainly saw that Expedia launched a special app that actually relies on ChatGPT to be able to recommend holidays and it's great to be able to see something that customers can in, in, in interact with and really gives ideas I think for the travel industry. However, there are concerns over AI and, and the ethics around it. As, as we heard a few weeks ago, there was a letter that was consigned by the lights of Elon Musk and thousands of others demanding a pause in artificial intelligence research, which obviously has created a firestorm and they asked for a six-month hiatus on the development of these systems.
Ryan Haynes:
What I understand about that, that's about the, it's the focus on the deployment of large language models, which you referred to, and these models can predict the next word in a sequence given the proceeding words and the, the backbone of many of the conversational AI systems. However, the experts are arguing that these models are more powerful than ever before and they simply don't understand how they work, which is quite scary if if if the people behind developing them don't understand how they work. So what other sort of ethics do we have to be considered especially to make sure that you know, that it's under control and we know what is actually delivering for our customers and also I guess employees and, and the larger workforce.
Emma Taylor:
I think yeah, there's a lot to unpack there. That's so interesting. I mean I think with the, with the letter from, from the, from the different experts, that was a really interesting thing because I've kind of perceived it that the, it's not the AI itself, which is getting out of control, it's more the arms race over AI. You know, they're sort of like one-upping each other constantly, which, which is getting everyone into a frenzy. I think that's what they need to put the brakes on. It might not be the models themselves because I think one of the people did describe it as, you know, compare it to human intelligence, which is still incredibly, incredibly far, far away from.
Emma Taylor:
But yeah, I think it's, it's definitely right to stop and think before implementing all this type of technology and really knowing, you know, the dangers that come, the potential dangers that come with it and making sure that you've explored that. So, you know, I, I said in my presentation, you need to think about the dangers of misinformation. These, these models can often conflate fake news with news fact and fiction and then produce that at a really alarming level. You've got to think about things like data privacy. So, there's been a lot of different articles recently about the data that you then put into these models where, where it's then replicated again, especially for the hotel industry, it has a lot of access to really personal, personal data in terms of copyright as well.
Emma Taylor:
And also, things like cybercrime. There's a lot of things that need to be considered before sort of, yeah, rushing to implement these types of things. And I think there are a lot of considerations as well for general AI. So, in terms of your data inputs, that's always something that you need to really make sure you've, you've considered your training data, make sure it's diverse, make sure it's large enough, make sure it's representative enough is really important to really define your outcome, make sure your, your models aligned. Yeah, there's a lot of, a lot of different considerations.
Ryan Haynes:
So, for the hospitality industry, as you say, there's a lot there to unpack, but as there are some really quick wins that you can put into place. Now the industry as a whole, as we've just covered, it sits on a huge amount of data and there's a lot of discussion really about, you know, around the sustainability of this data and the impact that, that it could have on the carbon footprint as well as on security. And this is an area that you specialise in Nick and, and you've been looking at.
Nick Wyatt:
Yeah, it is and I think it's always fascinated me the sort of travel industry's relationship with sustainability because the, there's always this slightly uneasy marriage of these sort of two ideas because there are some parts of the industry which just, you know, inherently create a carbon footprint, but there is, there are a lot of things that you can do, you know, to also minimize the impact. And I think the, again, the industry has made great strides, but this is what, this is a really fascinating one to me with the data centres and it's something you spoke about actually, Ryan, in, in this kind of wrap up sort of key takeaways session at, at the conference was around this sort of how data's the new oil and it is in the sense that it can be damaging if we're not careful.
Nick Wyatt:
You know, that was sort of one of your comments and it's interesting because you have these huge data centres that are processing all of this information, all of this data, they generate huge amounts of heat. They take lots and lots of energy to run and then it's kind of, well how is that energy being generated? Is it clean energy? Is it fossil fuel energy? You know, those things have to be considered. But then also what, what I took away from the conference was that there are now some sort of quite ingenious solutions coming up, how you can actually then harness the heat that these centres produce,
Ryan Haynes:
Who knows what's next? I mean, we have moved so fast in the last decade with technology, and I know there's a lot of catching up to do even from our own perspectives of what that's capable of. And where, for you personally, in your own personal life, do you benefit most, do you think from ai? Because, you know, we've, we've seen it in different entertainment platforms and different payment platforms, but for you, you know, where do you get the most benefit or the most enjoyment?
Emma Taylor:
Oh gosh, that's a really difficult question. Probably in a way that I'm not even aware of, to be honest. It's just used, you know, so perfectly in everything that we do. I think probably a lot of in-entertainment recommender algorithms I think are just so astute now. They're really good at recommending exactly what you need and when you need it. Also good at recommending things that you want and probably don't need. So, I'd say that's probably detrimental to my bank balance. But yeah, maybe, maybe Netflix, who knows.
Ryan Haynes:
How about yourself Nick?
Nick Wyatt:
Yeah, I would say similar actually. I think it is the personalization aspect, having sort of things that are recommended to me that do kind of hit the spot of what I, what my interest area is essential. And I've, I've had this in all watch, just yes, you've got streaming platforms, you know, music with things like Spotify, also with things like Netflix or Disney plus, you know, you get it there. But actually, the one that interests me is when you mentioned the ChatGPT Expedia example because OTAs I think are actually quite good at this. They, they do a lot with this, and they know what to push to you. They know that you know, oh, this person has booked on mobile devices before, let's push them a mobile-only offer. You know, like you get on something like booking.com for example.
Nick Wyatt:
So, I actually would say that, yeah, that kind of recommended piece. But to me, I've very much got a keen eye on that sort of, that planning stage of a trip use of large language models I think is really interesting because of the way that Expedia explains it, if I remember correctly when I read their statement when they launched this, is that they actually describe it as you can have an open-ended conversation with the platform, but then, you know, that's, that sounds really great, but that brings in your point then Ryan, about this, the diversity element. You need to make sure it's not leaning on misconceptions, mistruths, you know, kind of lazy tropes about certain destinations, certain attractions, et cetera.
Nick Wyatt:
Because that then risks becoming a negative and steering people away from those areas based on not a lot, but yeah, for me, I would say it's the recommended platforms and, and that to me, as I say, I think OTAs are, are an area I've definitely got an eye on with this. Because I think it, it could be fascinating to see what they do.
Ryan Haynes:
Yeah, I guess you got to make it fun, haven't you? I mean I saw last year that an OTA launched a, like a Tinder-style matching holiday matchmaker and it's based on your intent and preferences, and it was just, it was just grin, swipe left, swipe right. Oh my god, this is everything that I'm particularly interested in.
Nick Wyatt:
Yeah, so, but it's, it's interesting because I think that came up at the conference, didn't it? People were talking about gamifying certain things to make to cause it, it aids engagement with the platform essentially. But I guess the final piece where maybe I would like to see more done that came up at the conference, which I thought was really interesting, was on that sustainability piece and how, you know, this has gone on for years. You know, you get the, the sort of sign in the hotel room, oh you know, you help out a lot if you don't, you know, have your towels washed every day and all this kind of thing. But is there a way of tying that to loyalty schemes, for example, is there a way of in how can you actually properly incentivize the guest to engage in those things, to engage in those sustainability measures?
Nick Wyatt:
It could be anything from water usage to food wastage, but I think that idea of maybe linking it to some sort of reward to them, which could be loyalty points, I think is, is an interesting one. And it will again, you know, be, be interesting to see whether they go down that road of using AI to start maybe then to link it to that sort of thing.
Ryan Haynes:
Yeah, I mean, as you say, it'll be interesting to see all those use cases that start to come out and I'd definitely be looking at, look at exploring those here on travel market life. So, Emma, Nick, thank you ever so much for your time today and actually giving us insights into how the industry is using AI and some of the, you know, challenges to watch out for as well as the ethical dilemmas. It's been insightful.
Nick Wyatt:
Thank you
Emma Taylor:
So much for having us.
Ryan Haynes:
No problem. Hope to speak to you soon. So that was Emma Taylor, the thematic analyst to Global Data, along with her colleague Emma Taylor, Head of Research, talking about AI in travel, hospitality, and the ethics behind it. For more podcast episodes, check out Travel Market Life across all the podcast channels. Subscribe to our newsletter on LinkedIn. Thanks for listening. And look out for our next episode, ciao for now.
Comments