In 2002, I started my first blog. People thought it was odd to put your thoughts out on the Internet for all to see. I was speaking on panels, helping to define blogging policies for Microsoft, and my site traffic blew up. I also burned out on blogging and stopped after a while. I took up barbecuing and grilling as a creative outlet. Over time, I’ve taken classes from World Champions, became a certified BBQ judge on three circuits, co-founded the first Big Green Egg festival in the PNW. As a result, I started a blog, BBQHawks.com to share recipes with friends. Then, a few years ago, I stopped posting my experiments as life got in the way. Then OpenAI’s ChatGPT was released…and AI brought my joy back to BBQ and blogging.
It all started with a question:
Give me a great recipe for roasted chicken on the Traeger smoker.
The response was instant, with bulleted ingredients and numbered instructions:
I cross-checked details such as smoker temperature. and the safe internal temp for the chicken was spot-on (because food safety matters). Everything checked out. So, I went to work, following the recipe dutifully, taking pictures as I went along. The recipe turned out great. My wife loved it, and over dinner, we talked about how I should post it on the site. So, back to ChatGPT I went:
Generate a good blog outline for a post with a Roasted Chicken Recipe on a Traeger.
The result was a well-structured outline with 1) Introduction, 2) Materials Needed, 3) Steps to prepare, and then 4) Cook the chicken. It even provided suggestions for serving and storing the leftovers! The conclusion reminded me to summarize and include a call to action to try the recipe and suggest other recipes. I made my edits, adding the vegetables I cooked alongside the chicken. With AI, always keep the human in the middle.
Now I was speeding along in my workflow:
What title should I use for a recipe for Roasted Chicken Recipe on a Traeger?
The response was immediate: “Juicy and Flavorful: The Ultimate Roasted Chicken Recipe on a Traeger.” Not exactly mind-blowing, but it was clear and inviting, with an almost innate understanding of words that would grab the audience and search hits based on popular keywords.
Next up was search engine optimization:
What are the best keywords for a recipe for Roasted Chicken Recipe on a Traeger?
The results came back with fifteen numbered recommendations. It recommended the typical “Roasted Chicken” and “Chicken recipe” examples, as well as others I wouldn’t have thought of: “Poultry cooking” and “Juicy Chicken”. It seemed to know that a Traeger was a type of grill with its own audience, recommending “Traeger grill recipe” and “Traeger” as additional keywords/keyphrases.
The experiment went on. ChatGPT gave me a solid focus keyphrase, “Smoked and Roasted Chicken Recipe,” for further search optimization.
This entire process took maybe five to ten minutes. I copied and pasted the content into WordPress, added pictures, previewed, and published.
The Instagram Experiment
Now I was rolling. I created a Pinterest card via Canva which took only a few minutes and published. Then I went to work on an accompanying video reel on Instagram. Before I posted, I paused and asked ChatGPT:
What Instagram tags should I use for a recipe titled “Juicy and Flavorful: The Ultimate Roasted Chicken Recipe on a Traeger”?
The service started spitting out hashtags – 10, 20, 30, 40, 51 until I finally told it to stop. Too much. I know from experience that Instagram’s algorithms punish you if you have too many hashtags. Lots of paid-for mobile apps and experts claim to offer the best recommendations, so I asked ChatGPT:
What is the optimal number of hashtags to put on an Instagram post?
ChatGPT came back with the range of optimal number of hashtags, referencing “experts” as her source. It explained too few might limit reach, while too many will make my post look “spammy.” It summarized that it’s best to experiment and remember that Instagram allows a max of thirty hashtags. I hit publish on my Instagram Reel and added ten recommended hashtags.
Now, I have yet to see an overnight explosion in the popularity of my little blog. Pinterest engagement is up 6%, Insta is up about 10%, and I’ve seen a 15% increase in visits to the Blog. Not earth-shattering, and I won’t be leaving my job to become the next Bobby Flay. The real value so far is that it more than halved the time it took to publish. It removed the tedium of having to build the framework of the post and find the right contextually aware keywords/keyphrases and hash tags. No wonder publishing outlets, including The Guardian and Buzzfeed are now licensing ChatGPT to aid in writing articles. This is disruptive. Buzzfeed’s stock jumped over 200% on the news.
Of course, there are many professional and societal implications to consider from my little BBQ blog experiment. Will journalists be supplanted or augmented by AI? How will newly-graduated journalists hone their craft? What are the ethical implications – should AI-assisted posts include disclaimers, such as CNET was forced to add after being caught turning the once-storied site into an “AI-powered SEO money machine.”
The reality is every domain from creative to professional, and academic will have to grapple with the implications. Last week, Amazon warned its employees not to use AI tools to improve their business writing or do code reviews due to the risk of leaking company secrets. Educators are debating whether to embrace AI technology while ensuring academic integrity. Will a new low-code/no-code set of skills be required to generate effective prompts and need to be taught at universities? That was the question a professor at The University of Tulsa* texted me last weekend, just as I hit “publish” on my AI-assisted BBQ Blog.
Of course, this post begs the question – was this post generated by AI?
No. However, I did use Grammarly to clean it up.
Does that count? 🙂