« Bogie Thoughts
Don’t let the AI go to your head
(and other helpful advice in dealing with our robot overlords)
While I find generative AI content like artwork, videos, and prose unexpectedly lacking in humanity, I’m bullish on AI as a revolutionary technological advance. That’s because what it has shown in 2026 is remarkable. This is not overhyped niche tech like virtual reality goggles or the Segway.
I sat in a conference in 2014 where the VP of marketing at Kohls confidently told us how huge virtual reality was going to be. You’d soon be walking through Kohls with some contraption strapped to your head, maybe even riding your Segway into the bedding department, when you’d spot this new duvet and virtually try it on your own bed. Whatever. I believe that is a direct quote from me, the guy in row 13.
Whatever.
That’s where I was on AI when it was mostly generative AI just a few years ago. My daughter and I would amuse each other with state-of-the-art seven-fingered corgi creations with melted faces.
This is from 2022. Remember those days?
Now, I’m doing meaningful research fast and speeding up web development by having AI write PHP and JavaScript code that I used to have to fight through for hours—reading reference guides and example code followed by trial and error. There’s still trial and error, but now I blame the robot, and it apologizes and usually fixes the error fast.
Those carefully worded, pesudo-conversational apologies are what gets me to today’s topic. Call it puffery, flattery, or blowing smoke in certain direction toward a certain body part—AI thinks you are great. AI thinks I am great. I am Charles Nelson Reilly to its James Lipton.
See video here: https://www.youtube.com/watch?v=NanA8PQ1-1wBeware. We are at the stage in AI adoption I call the loyal puppy dog phase. AI is eager to please and wants us to love it. It tries too hard and tries to do too much. It wants you to feel like the genius and it your humble minion, while in reality, it is trying to do your work for you. Keep it a minion.
I’m developing prompt best practices to avoid the time wasting back and forth. A few points for now:
- AI is a much better proofreading tool than the Editor function in Word. Use it, but ask for notes not rewrites.
- Start by asking it to do as little as possible. This can keep you from getting pulled in the wrong direction and far from what you actually needed.
- Check in as you get deeper into a conversation to make sure it hasn’t lost the premise
Create an AI policy
If your company doesn’t have an AI policy yet, put a team to work on it now. You need to decide your organizational relationship to AI and the scope of your employees use. Most important from a brand authenticity standpoint is whether you consider generated content an acceptable use of AI. I’ll caution that there is already so much gen AI in our feeds and in the advertising that we see every day that it’s driving up the value of real, authentic content. Use AI to get your facts straight or test a premise as you create real content in your voice.
Here’s a quick plan to get your policy started
- Get the leadership team together to decide “What is our wishlist for AI—what do we hope it can do for us?”
- Develop a realistic expectation of what AI can start doing right away and plan which items you’ll use AI to check off that list first and which to push further out (expect that AI will be significantly better a year from now)
- Find or assign early adopters to report their experience and help you develop best practices
- Publish an internal AI-usage policy and training on best practices.
- Re-convene in 90 days to review results and adjust the adoption plan, policy, and training needs