After so much buzz about AI over the past two years, things have grown a bit quiet.
Was it all just marketing bluster, or is generative AI still poised to transform how we work and live?
For hoteliers, is it enough to dabble occasionally with ChatGPT or Gemini, or do you actually have to learn how to use these tools?
To find answers to these questions, it’s helpful to look at the path new technologies typically follow from inception to mass adoption.
Charting AI on the Hype Cycle
Decades ago, Gartner introduced the Hype Cycle to map the maturation process of emerging technologies. The model is intended to help people and companies separate the hype from commercial viability and avoid adopting technology too early or too late, giving up too soon, or hanging on too long.
The model identifies five key phases:
- Innovation Trigger – New technology breaks through with a product launch, and startups and venture capitalists jump on the bandwagon.
- Peak of Inflated Expectations – Excitement attracts more suppliers and users, generating media coverage and buzz.
- Trough of Disillusionment – The excitement wears off, as early adopters report performance issues and low returns on investment.
- Slope of Enlightenment – Early adopters begin to see initial benefits, while others start to understand how to adapt the innovation to their organizations.
- Plateau of Productivity – A growing number of users recognize the real-world benefits, and the technology goes mainstream.
Charted on a graph, the Hype Cycle looks like a rollercoaster—thrilling for some of us and terrifying for others. The question is, where are we in the AI Hype Cycle today?
Careening Up the Peak of Inflated Expectations
When ChatGPT debuted in November 2022, it attracted over 100 million users in under two months, the fastest-growing consumer application in history. By comparison, it took the internet seven years to reach that number and Facebook and YouTube over four years.
Shortly after, Google, Anthropic, Meta, and others launched their own large language models (LLMs), and users flocked to them too.
As early adopters discovered the “magic” of AI, many predicted it would “revolutionize” virtually every field of work and study, help solve the world’s biggest problems, and possibly even lead to our destruction.
By February of this year, 23% of U.S. adults reported they had used ChatGPT, up from 18% in July 2023, according to Pew Research.
Idling in the Trough of Disillusionment
The problem is a large proportion of users play with LLMs once or twice, maybe think they’re cool, but then don’t go back.
Why? There are numerous reasons, including:
- They didn’t see how it was useful to them.
- It couldn’t do what they wanted.
- The results were unimpressive.
- They don’t have time or are unwilling to learn new tools.
- It wasn’t required as part of their work.
One of the biggest turn-offs is the tendency of LLMs to give incorrect information and hallucinate (make stuff up). Apparently, this is a feature, not a bug.
So, does the journey end here? Will genAI join the ranks of overhyped tech like voice assistants, Google Glass, virtual reality, the blockchain, and the metaverse? (Although some of these technologies are reemerging in different forms.)
Scaling the Slope of Enlightenment
On the contrary, I think we’re experiencing the calm before the storm. Behind the scenes, tech companies are quietly developing AI technology, and organizations are figuring out how to integrate AI tools in the workplace.
Early this year, a survey from Bain & Company found that 87% of companies were developing, piloting, or have deployed generative AI in some capacity.
In the hotel industry, AI-powered chatbots are popping up everywhere, and early AI use cases are piling up. (Check out the roundups from Ira Vouk in HFTP Tech & Fin Hospitality, Douglas Rice in Hospitality Upgrade, and Adam Mogelonsky in The Hotel Yearbook.)
Developing AI technology takes time and resources. It’s not enough for a hotel to simply plug into an LLM and call it an AI chatbot. The outputs would be too generic and unreliable, and the risks of misinforming guests too high.
To dispense accurate, hotel-specific information, the LLM’s knowledge base must be supplemented with contextual information from an external data source like the hotel’s guest directory, PMS, or CRM.
Moreover, not all AI is generative AI. Standalone tools take even more time to develop, whether they’re AI-native applications or are integrated into existing software.
It also takes time for hotels to develop an AI strategy, research and vet AI solutions, and analyze the impact on the labor force. They must also address concerns about data security, privacy, and the responsible use of AI before implementing tools and onboarding employees.
But it’s happening. That silence we hear is the sound of the groundwork being laid.
Next Stop: the Plateau of Productivity
In his book Co-Intelligence, Ethan Mollick cites various studies that have shown that AI can bring improvements in productivity of 20 to 80% across a broad variety of job types. And it’s only getting better. Current versions are “the worst AI you will ever use,” he points out.
Mollick believes AI will evolve into General Purpose Technology—once-in-a-generation technology that touches every industry and aspect of life, like steam power or the internet.
In some ways, it might even be bigger. “We have invented technologies that boost our physical capabilities and automate complex tasks, but we have never built a generally applicable technology that can boost our intelligence,” he writes.
As AI technology overcomes its limitations (or finds workarounds) and more users and companies integrate AI tools into their workflows, genAI is poised to go mainstream.
For hotels, AI might even bring the breakthrough we’ve been waiting for: an intelligent interface that connects to all software in a hotel’s tech stack and serves as an all-knowing, conversational gateway to the property for guests and staff alike.
The AI rollercoaster ride is far from over. If you want to join the ride, you had better buckle up—and buckle down by learning the tools.
