AI as job security for the industry

Date – Wednesday, 25 September
Time – 9:30 – 10:30
Location – Auditorium 6/7, CCL

AI might be leading us into a golden age of travel – giving us more free time and making just about everything more efficient. It will also make travel plans harder to trust and human knowledge and assistance from travel professionals more valuable than you think.

A few years ago big travel brands threw 6-7 figures into creating AI concierges, many of which weren’t really AI, and served mainly to displace call centers or drive upsell / attachment. Fast forward, and mass democratization of AI means any BnB can publish a GPT for $20/month. The result is a flood of AI-generated or AI-intermediated content and experiences all promising to be the perfect guide, pretending to care about user preferences with five-point quizzes to plan your perfect travel.

This is simultaneously empowering for travelers, who employ their own AI-directed planning tools, but who feel an illusion of personalization, when in fact they’re getting the same set of “off the beaten path” recommendations in response to their prompt as everyone else. Why? Because since November 2022 we’ve been living in a land of foundational models which don’t know their users.

That’s all about to change. Restaurant recommendations are a lot more powerful when they know every reservation you’ve ever made on the internet, every place you took an Uber to, or every review you posted on Yelp or Google Maps. The new round of agents will start from a totally different place. If you didn’t pay cash, you’re training the algorithm. The good news is that the personalized can now plausibly be personal; the bad news is that you as a traveler won’t actually know whether the recommendations are organic or paid, driven by partner programs and sponsorship.

In other words, your guide may be taking bribes.

Then comes agentic AI. Once your AI tour guide has your credit card, your prompt can kick off a series of bookings – “I’m feeling lucky” results in not a search result, but three days planned in Tulum. Risks include hallucination which no longer just invents restaurants, but addresses and the names of tour guides. Safety measures will be introduced to validate plans and destinations, but first we’re likely to see social media filled with stories of people who ended up on a dirt road searching for ancient ruins and found only a flash mob of other travelers who were pranked by an errant post, ingested into a foundational model.

AI is really good at intersectional thinking. Looking for a gay-friendly volunteering trip? How about an eco-tourism conference destination? What previously might have been niches deemed small for content creation or dedicated offerings will be supported by AI supply and demand.

Misinformation and disinformation campaigns will become common – if a competitor or bad actor seeks to do economic damage to a travel destination, then seeding of models with horrible content about crime, pollution, or poverty will do the trick. Algorithms will struggle to separate the malicious from the authentic experiences of travelers.

Meanwhile, MR/VR staycations will sound a lot more appealing when the world is melting in heat waves and regional conflicts – tourism in Vietnam will not benefit from increasing tensions in the South China Sea. With AI-generated video, experiences will cross the uncanny valley and become closer to indistinguishable visually.

Increasing effects of climate change will drive a binge of tourism to endangered locations followed by a famine. Concerned about AMOC collapse? Now’s your chance to visit Iceland! AI will fan the flames of this, amplifying trends and divisive content in climate FOMO.

At the same time, we will hear more about a golden advent of leisure time from those who expect AI to provide it – what would the impact be on travel? Let’s explore it together in Lisbon.