It’s everywhere, or so it seems. Artificial intelligence has rapidly become part of the media and information landscape, and a problem-solving aid.
Komoot is the latest to jump on this bandwagon. It revealed an app on Wednesday that uses ChatGPT “to discover and plan outdoor routes using natural language prompts”.
The app enables you to request routes by location, activity type, length and more “simply by typing a question”. It then draws upon Komoot’s library of 7 million routes and 4 million highlights, with suggestions that are more relevant to your request than if you were to use Komoot's standard app. These suggestions can then be imported to Komoot for editing, saving and offline use.
But do we really need it? I’m not so sure.
ChatGPT’s ascendancy
In the three and a half years since OpenAI launched ChatGPT, it has become a global phenomenon. OpenAI reported in September 2025 that it had 700 million weekly users.
OpenAI said people use the AI chatbot primarily for “getting tasks done”, with three-quarters of conversations focusing on “practical guidance, seeking information, and writing”.
Meanwhile, Adobe ran a survey last year in the USA that revealed 77 per cent of people use ChatGPT primarily as a search engine, with 24 per cent going to ChatGPT before other platforms.
This is part of the rationale behind Komoot’s new app, according to project lead Felix Tarcomnicu: “As people increasingly use ChatGPT for everything from daily tasks to trip planning, Komoot’s app brings Komoot directly into the workflows people are already using to find inspiration and shape their plans.”
“This evolution of Komoot’s navigation and route planning technology will make exploring the outdoors even easier for people worldwide,” Tarcomnicu adds.
A frictionless existence
Komoot sums the app up by saying it “removes friction from adventure planning”. This idea of removing friction lies at the heart of the sales pitch behind AI platforms such as ChatGPT. Yet it strikes me as totally at odds with life in the outdoors. Sure, we want to find great places to explore. But where’s the exploration in a frictionless landscape?
The debate on how tech companies have infiltrated daily life led journalist Kathryn Jezer-Morton to cook up the phrase “friction-maxxing”.
In The Cut, Jezer-Morton advocates for avoiding things that supposedly make your life more convenient. She says technologies such as ChatGPT, and smartphones more generally, provide a form of escape from difficulty and thereby infantilise us. By making things so easy, they rob us of the satisfaction of putting the work in to figure things out.
Jezer-Morton says friction is “really the only defense we have against the life-annihilating suction of technologies of escape”.
Resisting the friction-free experience of an AI route-planning app therefore feels to me like the most certain way to reclaim escape.
There is a joke on the internet that people who spend too much time online should “touch grass” to reconnect with the real world. So maybe next time you feel this desire, you should leave your phone at home and step outside to find your own patch of sweet, green grass.
And hey, maybe you should use a paper map and a compass. The friction of orienteering is far more satisfying than following a dot on a screen, after all.
But wait, there’s more
I’m not naive to the fact that we already use convenient, online tools to build or decide on routes. Scouring Komoot has unlocked many great cycle rides for me. But there’s a deeper issue with the use of ChatGPT to access the outdoors. The environment.
Karen Hao’s book The Empire of AI plots the rise of OpenAI. Hao has arguably had more access to OpenAI than any other journalist, and first covered the company set up by Sam Altman back in 2019. Her book tracks how the tech company has exploited humans and the natural environment around the world to build that AI tool in your pocket, ChatGPT.
Generative AI models such as ChatGPT (GPT stands for 'generative pre-trainer transformer') rely on huge amounts of data to provide your prompts with answers. They therefore require a huge amount of ‘compute’ power from chips, and an alarming amount of water and energy.
According to Hao, the data centres used to power AI are entering a new phase, and scale. Developers call them megacampuses to reflect their size, but also the “sheer amount of energy” needed to run them.
Hao says that a rack of Graphic Processing Units used for AI requires three times more power than a rack of other computer chips. She says, according to the International Energy Agency, “each ChatGPT query is estimated to need on average about ten times more electricity than a typical search on Google”.
The largest data centres were 150-megawatt facilities, which could consume the energy required to power 122,000 American households. But Hao writes: “Developers and utility companies are now preparing for AI megacampuses that could soon require 1,000 to 2,000 megawatts of power. A single one could use as much energy per year as around one and a half to three and a half San Franciscos.”
By 2030, Hao writes, AI computing could use more energy than all of India, the world’s third-largest consumer of electricity. And by 2027, she says AI could consume 1.1 trillion to 1.7 trillion gallons of fresh water globally. Added to this, the hardware behind AI and data centres requires the extraction of resources such as copper and lithium.
Such demands have led tech companies to jettison their environmental commitments. Google revealed in July 2024 that its emissions had climbed nearly 50 per cent in five years due to AI energy demand. The AI push at Microsoft, which partnered with OpenAI in 2019, saw its carbon emissions grow, making its goal to be carbon negative by 2030 harder to reach.
This stuff might feel abstract and insurmountable when considering an app that’s designed to help you find routes based on simple prompts. But I struggle to square technology that leads to emissions growth and intense resource extraction with a love of the outdoors. I’m asking for a bit more friction, and maybe that’s something we’d all benefit from in the long run.
