It seems there is no corner of technology that is considered “safe” against the AI boom. Every popular piece of technology, from earbuds to laptops, uses artificial intelligence. Apples WWDC 2024 Developer Conferenceshowed up, for example Siri’s new ChatGPT partnership, part of the company’s new Apple Intelligence service.
Even the best fitness apps are not immune. First, the running and cycling app Strava made waves with its Athlete Intelligence AI analytics feature, and now it’s AllTrails’ turn. One of the best fitness apps for outdoor adventurers, AllTrails helps hikers find new trails (perfect for to get those 10,000 steps), which collects comments, photos and reviews from other hikers on approximately 400,000 trails.
AllTrails debuts with a redesigned Trail page and a handful of new features, which were presented to me during an exclusive conversation with James Graham, Chief Technical Officer of AllTrails. First up is a redesign of the individual Trail pages, which Graham says are “the most popular pages in the app.”
The distance and difficulty levels are now at the top of the page and you can use the new Photo Tour feature to see the key points along the way. AllTrails uses machine learning combined with “geoclustering” to look through the photos of each trail taken by users so that the photos appear at the right points. Along the way you can see exactly where people took certain photos. Smart.
You can also find reviews of the route on each route page, and AllTrails has worked with OpenAI on what Graham calls “ReviewGPT”: a ChatGPT-powered summary of all reviews on the route can now be seen at the top. every review section on the Trail page. For example, if you open a near me page, the following AI-generated summary will be displayed:
“Recent reviewers of Shiplake Circular… mention that the path can be muddy at points, making walking boots necessary. The route offers varied scenery and is recommended by visitors. Although some areas along the river are muddy, there is currently no flooding. It is suggested to walk the route in reverse order and climb the hill first.”
The app also offers a new collections feature, allowing you to sort local routes by categories such as ‘Under Trees’, ‘Epic Views’, ‘Wildflowers’ or ‘Dog Friendly’. AllTrails has also introduced extensive parks pages, with complete guides and information about national parks, including popular routes, rates, parking, visiting times and more.
Although the app is currently fairly focused on the US due to the large selection of national parks (and the app makers are of course based in Silicon Valley), there are guides to parks outside the US. In Great Britain there are guides to ten parks in England, two in Scotland and three in Wales.
Analytics: Useful AI features, but what’s the security like?
While it makes a lot of sense to use AI in this context (no more scrolling through reviews to find out what gear to bring and whether a trail is dog-friendly), by pouring information about our location, health and walking habits into a third . party AI service could be a cause for concern for skeptics.
Graham quickly assured me that AllTrails user data was safe. “We have certain assurances that OpenAI will not train and use any user context we provide to train their models or use user data,” he said.
“We also use RAG – Retrieval Augmented Generation – which allows the LLM to request specific data so that it never actually ends up in the LLM. Data segregation exists at the architectural level to ensure user privacy is guaranteed.”
Graham later shared the following on behalf of AllTrails:
“To ensure that our user data remains secure, we adhere to the OpenAI Data Sharing Agreement, which guarantees that data sent to the OpenAI API will not be used to train or improve OpenAI models .Our pre-trained model usage remains within our secure AWS environment and is not shared, sold or accessed by third parties.”