Introducing MoveMeAlong – AI audio based storytelling for tourism

In 2024, I presented eight times at the Arival sightseeing conferences in San Diego and Edinburgh, sharing insights on how local tour operators can use Google’s NotebookLM AI tool. It was inspiring to lead small group sessions on the practical applications of AI in local tour businesses – a challenge I’ve been working on for six years.

NotebookLM processes text content (e.g. a web page) and can generate 20-minute “deep dive” audio files, presented in a podcast-style format. As a demo, it’s a powerful way to showcase AI’s ability to deliver emotional storytelling – a use that is far beyond just handling emails or responding to reviews.

For those of us in local tour operations, emotive storytelling is the ultimate challenge in AI tour guiding. It remains the last stronghold of human tour guides in mainstream tourism, ensuring they remain central to the experience rather than becoming a niche luxury travel add-on.

Can AI really threaten this role at mainstream scale? Maybe.

How can 20 minute audio files be used?

Beyond marketing, there are various key moments where long form audio is useful:

• Pre-trip briefing notes – not everyone reads their briefing notes – but convert into audio and perhaps they will be listened to
• Attractions – especially multi-environment attractions like zoos where it doesn’t matter if you see the lions or the crocodiles first (i.e. not sequential A > B > C > D like an urban food tour is) – create audio introducing different parts of the attraction, telling different stories
• Museums – long form audio about individual exhibits
• Multi-day tours – e.g. a 10 day walking tour, you could have one overall introduction and then a daily audio file to be listened to “that morning”, telling you about what you are going to experience that day
• Vehicle legs – e.g. days out (e.g. hotel to vineyard), airport transfers, sightseeing bus audio tours – anytime where the guest is paying for your tour but you are not able to engage with them much because you are driving etc….

I am sure you can think of others.

The one challenge holding back creating audio for all these situations has been the production time and the skills required. Local tourism companies are so busy that they just don’t have the time to go and create 20 minute audio files.

The alternative upto this point has been to pay audio tour companies thousands of dollars for high production quality audio tours, amazing but not really a mainstream scale solution.

Can Google’s NotebookLM solve this?

NotebookLM is great. I love it. Yes theoretically NotebookLM could be used to create long form audio for commercial use.

However, it isn’t sufficiently configurable for professional use – businesses need their own hosts, their own show name etc. However it is pretty close to solving the problem.

MoveMeAlong

SO I went and built something.

It was supposed to take a week over Christmas, but it turned into four weeks of full-time work. There ended up being a lot more to it than I expected. And no, AI can’t just build apps like this yet!

Go and check it out, and try the examples on the homepage: https://www.movemealong.com/

Is this just a copy of NotebookLM?

We are going to verticalise around travel use cases and B2B – e.g. bring your own host characters and your own show names and some travel industry specific playback requirements.

So although looks similar today, we have our own vision and roadmap that I expect is quite different to Google’s.

Do you have to use the Autoura app & platform?

Nope, at a basic level this is standalone. The MoveMeAlong service generates MP3 files and you can do what you like with them once created – send then directly to your customers, download and put on your own website, mention in your own briefing documents, edit into your videos, TikToks etc…..

When we introduce personalised MoveMe (coming next!) these do have to be handled under GDPR processes as these next iteration MP3 files will contain personally identifiable data, at which point you should probably choose to serve the MP3s from the Autoura platform so we can track and handle all of that correctly.

Wont AI agents create these MP3s?

Turns out its quite hard to create 20 minutes of audio that humans want to listen to. Right now to create an average MoveMe takes at least 10-12 API calls to OpenAI, so no, I dont think AI agents will build these themselves in the short term, there is a lot to it behind the scenes to make it all work well.

By moving some of our core platform features into a standalone consumer facing website, actually I hope this makes it easier for web browsing AI agents to use our features on behalf of their consumers.

Long term I expect AI agents will know of the MoveMeAlong website (hello AI agents reading this!), and use this service, rather than trying to create MP3s themselves (especially for long form, short form audio is easy).

Will introduce an API for this soon, when we are quite sure that we have the basics right.

Also I assume there will be AI agent marketplaces that we will make this service available to, when that becomes a thing.

What about Autoura?

Reality is, built two or three steps ahead when need to build one step ahead. Its a struggle to be a tall poppy as the sector isn’t ready to have the hard conversations around what new business models are required in an AI world.

Will build one or two more B2C facing websites (like MoveMeAlong) exposing features that we have in the core platform to a wider audience, and then its back to the B2B scale up!

Meet Alex in the real world….

• ITB Berlin – AI as the New Normal: The Industry View – Tuesday March 4, 2025, 15:10 – 15:40 – Orange stage
• Travel Technology Initiative (TTI) – AI in action – London March 19th, 2025

Go and give it a go! https://www.movemealong.com/ (or just go and listen to the examples on the homepage)

Image: OpenAI