Earlier this month, three novice climbers headed up Mount Shasta after using AI as part of their planning process. They reportedly used Google Gemini for advice about the route and what provisions to pack. Unfortunately, they got lost, ran out of food and water, and ended up spending the night on the mountain before being rescued. This isn’t an isolated instance. Rangers around the country have reported an increase in problems due to trail users relying on AI information for the backcountry.
The problem is not necessarily using AI. Controversial as it is, nowadays it’s almost impossible not to use AI in some way if you use the internet. But problems can occur when users treat AI like a subject matter expert or a verified source.
How Does AI Work?
When you ask an AI chatbot a question about the backcountry, it isn’t pulling a trail guide off a shelf and looking up the answer. Artificial Intelligence (also called “large language models” or LLMs) comes up with answers by predicting language based on patterns from existing information from the Internet and other sources. What exactly does that mean?
Think of AI like an extremely sophisticated autocomplete. It has learned patterns from enormous amounts of information and uses those patterns to construct an answer to your question. If, for example, you ask: “Where can I park my horse trailer to ride the Imaginary Creek Trail outside Ely?”, AI may recognize the pattern of questions about trailheads and Nevada recreation and produce a perfectly plausible answer about a parking area, road access, mileage, etc. But if reliable information about that trail isn’t available, or if Imaginary Creek Trail doesn’t even exist, some AI systems may still generate an answer rather than simply saying, “I don’t know.”
AI is extremely good at producing information that sounds coherent. However, a convincing answer and a verified answer are two different things. Even if the source material where the underlying information came from is incomplete, outdated, ambiguous, or flat-out wrong, the resulting answer can still sound completely confident.
Why this Matters in the Backcountry
Using verified information is incredibly important when it comes to recreating safely in the backcountry, especially in Nevada. AI has much better information about Yellowstone or Yosemite than it does about some obscure Nevada two-track leading into a mountain range 60 miles from the nearest town. Why?
Because there are simply more sources about Yellowstone and Yosemite to pull from. Some obscure Nevada trail may have very little information online, old or conflicting information, outdated maps, routes that have changed, washouts, regulations that may have changed, and so on. AI cannot magically know information that doesn’t exist or isn’t reliably documented in its available sources.
Using generative AI to create maps often produces trails that don’t exist, connections between two trails that don’t exist, incorrect distances, fictional trailheads, bad coordinates, roads that aren’t passable, etc. I’ve now seen AI-generated Nevada trail maps shared online that contain exactly this kind of problem. Never use an AI-generated map as your primary navigation map!
Where AI Is Useful for Trip Planning
If you use AI, you have to think of it as one of your tools – not as an omniscient being. It doesn’t just automatically know everything, and it doesn’t have judgment. Here are some ways you can integrate AI into your research:
Brainstorming: “I want to take my horse camping for the first time. What should I think about when choosing a destination and preparing my horse for the trip?”
Building checklists: “Help me make a packing checklist for a three-day horse camping trip where I’ll be hauling my own water and feed.”
Organizing information you’ve already gathered: “Here are the horse regulations, campground rules, and wilderness regulations I found on the Forest Service website. Organize them into a simple reference sheet for my trip.”
Thinking through contingencies: “I’m planning a remote day ride with my horse. Help me think through what could go wrong and what I could carry or plan ahead to deal with those situations.”
Generating questions: “I’m calling the ranger district before hauling horses to a trailhead I’ve never visited. What questions should I ask about trailer access, parking, water, current trail conditions, and horse regulations?”
Don’t Give Up on Primary Sources
Primary sources are original, first-hand accounts created by people who directly experienced or witnessed something. When you are researching new trails or planning a backcountry trip, primary sources that should be consulted include:
- Official land-management agency information (BLM, Forest Service, State Parks, etc.)
- Current topographic/navigation maps
- Current weather and fire information
- Recent firsthand trail reports
- Local equestrian organizations/resources
- Nevada Horse Trails listings
- A phone call when something important remains unclear
And don’t forget to cross-reference information. One website saying something doesn’t necessarily make it true. One AI saying it definitely doesn’t! For important details like water, access, closures, regulations, and trailer access, verify them through a primary source.
This is why Nevada Horse Trails puts so much emphasis on firsthand information. When we tell you what a trail or a parking area looks like, that’s because we physically went there. We hauled a trailer in. We looked at the parking. We rode the trail. We photographed the things. We researched the regulations. We wrote the listing ourselves.
AI Is a Tool. Treat It Like One.
Technology has changed outdoor recreation for the better. GPS has made backcountry navigation easier. Satellite communicators have made the wilderness safer. Digital maps make researching routes easier than it has ever been. AI can be another useful tool for people planning adventures. But none of those tools replace sound judgment, which comes from lived experience.
