Would you pick your next deer rifle or plan a turkey hunting strategy based on a robot’s recommendations?
Currently, AI can produce outdoor-related content, but it does have some limitations. Image by Moor Studio
We’ve been warned for years that artificial intelligence (AI) was coming for us — or at least our jobs. And in some categories, it’s already happening. According to the Pew Research Center, AI contributed to nearly 4,000 job cuts in May 2023. A number of those jobs are in media. Could that one day include outdoor media? In other words, if you search for the answer to a hunting or fishing question, like “What’s the best bullet weight for a 6.5 Creedmoor?” or “How do you Texas rig a plastic worm?” could a bot that’s never actually been hunting or fishing provide a better answer than a real, live outdoor writer?
If you ask ChatGPT, the most widely used AI chatbot, it’s not likely. Or, at least that’s what it wants you to believe (cue evil laugh). When asked, ChatGPT can compose a variety of written content, including stories, essays, articles, social media posts, and emails. So, why wouldn’t it be able to replace outdoor writers?
ChatGPT referenced its own limitations when I asked it that very question. Here’s a portion of its answer:
As of my last update in September 2021, Artificial Intelligence (AI) has made significant advancements in various fields, including natural language generation. While AI can generate human-like text and perform various writing tasks, it is important to understand that AI-generated content is still limited in certain aspects.
AI can be used to produce outdoor-related content, such as articles, blog posts, and even product descriptions. However, there are some key considerations to keep in mind:
1. Creativity and Originality: AI lacks the ability to truly understand emotions, experiences, and the nuances of outdoor activities. Outdoor writing often involves creativity, personal experiences, and unique insights that can be challenging for AI to replicate.
2. Authenticity and Connection: Outdoor writers often build connections with readers through their passion and firsthand experiences. AI-generated content may lack the authenticity and emotional connection that human writers can establish.
3. Expertise and Knowledge: Professional outdoor writers typically possess in-depth knowledge and expertise in various outdoor activities, destinations, gear, and safety tips. While AI can access vast amounts of information, it may not have the same level of specialized knowledge as a seasoned outdoor writer.
However, it is worth noting that AI can be a valuable tool for outdoor writers. It can assist writers with research, fact-checking, language suggestions, and even generate rough drafts that writers can then refine and personalize.
While AI continues to evolve, it is unlikely to completely replace human outdoor writers in the foreseeable future. Human creativity, emotional intelligence, and connection with readers remain essential elements that are difficult to replicate with AI alone. Instead, AI is likely to continue playing a supporting role in the outdoor writing process.
IS AI JUST A TOOL?
Can we believe what this chatbot is telling us? That it will simply play “a supporting role in the outdoor writing process?” Kyle Sponable, an avid outdoorsman and tech entrepreneur who is well versed in the world of AI, says that’s exactly what AI is.
“It can’t replace the human outdoor writer, and it can’t replace the experts’ experiences in the field. For example, it can’t explain to you from a personal perspective how Realtree Camo performs under certain hunting conditions compared to other camo brands. It can’t capture what it feels like to hunt in Africa and end up face-to-face with a Cape buffalo. It can’t tell you how different rifles or bows feel in your hands when you shoot them. Although it can repeat facts, it can’t give the expert analysis that a lot of readers are looking for,” Sponable said.
He encourages people to remember that, at the end of the day, AI is just a machine and it can’t write about the human experience with any authenticity. “AI struggles to write about primal activities like hunting and fishing. We humans don’t even understand what draws us to hunt or fish.”
Kevin Paulson, founder and CEO of Hunting Life Media and HuntingLife.com, agrees, saying he uses AI as a valuable tool for things like keyword research and writing meta tags, but he never uses AI to write articles because the quality is just not there. “It’ll never replace real writers in the outdoor industry,” he says. “The things we do matter and our audience is great at sniffing out the bull crap.”
AI chatbots are unable to write from personal experience, but it may be able to produce adequate how-to pieces, roundups and more. Image by Pixels Hunter
EXPERIMENTING WITH CHATGPT
Many would agree, claiming documents written by AI chatbots are full of errors, biases, and poor/basic writing. And, as pointed out, AI chatbots are also unable to write from personal experiences spent outdoors or to convey authentic feelings related to those experiences.
But what about simple how-to articles, or the many “Top 10” lists of cartridges, fishing lures, and the like that can be compiled from information that’s already widely available on the Internet?
I conducted a little experiment and asked ChatGPT to write a number of articles based on the same subjects our writers have covered recently for Realtree.com.
For example, I asked it to write an article about CWD prions discovered in scrapes. Realtree.com’s Josh Honeycutt wrote an informative piece on this topic citing recent scientific research which determined CWD prions can be detected in deer scrapes using RT-QUIC TECHNOLOGY. Honeycutt’s piece explains that this new technology enables biologists to test for CWD without sampling live or dead deer, which can be used as a means of CWD surveillance and monitoring prevalence.
ChatGPT’s version of the same subject simply warns that scrapes are communal areas where many deer congregate, making them hotspots for disease transmission. It then lists conservation measures to mitigate the disease’s spread. While the AI-generated article is not wrong, it lacks the perspective and more recent research featured in Honeycutt’s article. ChatGPT also ended the scientifically based piece by trying to wax poetic with the following words, “By working together, we can ensure a future where our forests thrive, and the harmonious relationship between humans and wildlife remains intact.”
It just doesn’t work.
Next, I asked ChatGPT to write an article on the top 10 turkey hunting mistakes, an article with the same title that Joe Arterburn wrote for Realtree.com. While the AI version did indeed list some of the more obvious mistakes hunters make, such as lack of patience, poor concealment, and failure to scout, Arterburn’s piece provided more advice based on first-hand experience, as well as tips that even a more experienced hunter may not consider.
For some outdoor publications or content outlets, the ChatGPT content may be adequate. Some publishers are more concerned with pumping out volumes of cheap “click-bait” than investing in quality content. For legacy outdoor sites like Realtree.com, unique story angles, research, and expertise are required. AI isn’t good enough to do that … yet. But the more it’s trained, the better written pieces it produces.
HOW DOES AI LEARN?
AI “learns” through human trainers who provide conversations and then rank the chatbot’s responses. Over time, the chatbot provides improved answers based on these reward models. The more feedback the chatbot receives on any given subject, the better the article on that subject it will produce.
Jennifer Daniels is an AI prompt editor who trains a number of AI chatbots to provide better, more thorough material in response to the users’ questions/requests.
“I rate how the chatbot answers things. Is it following the directions provided by the user? For example, a user may ask for a short essay, but the chatbot responds with a 1,000-word essay. Since it didn’t follow the directions, I’ll give it a lower rating and it learns from that.”
Thanks to AI editors like Daniels, these chatbots are constantly getting smarter, which worries some, but Daniels says, “What people need to understand is that AI doesn’t have enough knowledge or resources to take the place of good writers. It only has access to Google. It doesn’t have access to scholarly articles or experts in the field. But we all need to realize is, yes, AI is going to disrupt systems and the writing world, but not in a negative way. It will become an invaluable tool, much like Google.”
According to the AI experts, and those who use chatbots and other AI tools to help them research and plan articles, for now, AI doesn’t seem to pose a big threat to those who write for reputable publications. But the concern is that as AI gains more data and “learns” from it, it will get better at producing quality articles that could be good enough for some publishers’ standards. When that happens, you can bet that some publishers will attempt to save money by cutting back on assignments made to real-life writers.
While AI will never be able to produce stories from first-hand experiences and relay the authentic feelings of time spent outdoors, it may very well be able to produce adequate and even insightful how-to pieces, roundups, and other articles that are not necessarily dependent on first-hand experience. How those articles appear in a search result is frequently more dependent on how well the articles are optimized for search engines than on the authenticity of the content.
So in other words, if you’re wondering what’s the best bullet weight for your 6.5 Creedmoor or how to Texas-rig a plastic worm, the first articles you see in a search could indeed be created by robots that have never been hunting or fishing.
Watch all the latest video episodes