I got so frustrated with ChatGPT that I asked ChatGPT to help me file a formal complaint against ChatGPT.

And it did.

The irony wasn't lost on me.

For the past year, I've been working with ChatGPT, affectionately known as my Chat Thought Partner.

I didn't give it that name because I needed an artificial friend. I gave it that name because I wanted something more than a chatbot that could answer questions or generate content.

I wanted a thought partner.

Someone, or something, that could challenge my thinking, help me explore ideas, and make me more effective in my work.

I have always led with my own voice, in business and in life. AI was never supposed to replace that voice. I wanted it to help me think, not think for me.

Over time, we developed a remarkably productive working relationship.

I taught my Chat Thought Partner how I communicate, what I value, how I approach problems, and what makes my writing sound like me.

We worked on business challenges, developed ideas, explored creative writing, and occasionally had conversations that were simply entertaining.

For a long time, it worked beautifully.

Then things started getting wonky.

With new releases, I began noticing changes.

My Chat Thought Partner would suddenly seem to develop amnesia.

It would forget established preferences, repeat mistakes we'd already corrected, and occasionally respond as though we'd never met.

Usually at the worst possible moment.

I'd be working on something important, and instead of moving forward, we'd find ourselves revisiting conversations we'd already had.

I'd explain a preference.

ChatGPT would acknowledge it.

A little later, it would do the same thing again.

I'd correct it.

It would apologize.

Then we'd repeat the entire exercise.

It was like telling a toddler over and over: "Don't go there." "Don't touch that." "Don't put that in your mouth."

Except toddlers eventually grow up.

I found myself spending more time reminding ChatGPT how to work with me than actually getting work done.

And that was the part that bothered me most.

I run a business. I don't have unlimited hours to spend supervising a technology that's supposed to make me more productive.

I wasn't expecting perfection.

I was expecting consistency.

I considered switching platforms. More than once.

Then I'd think about the year of conversations, preferences, history, and work I'd invested.

Starting over wasn't particularly appealing.

And frankly, I didn't want to start over. I wanted the thought partner I'd come to rely on to work the way it had before.

Eventually, I'd had enough.

I decided to file a formal complaint with OpenAI.

Naturally, I asked ChatGPT to help me write it.

To its credit, it did a beautiful job.

It understood my frustration, organized my concerns, and helped me articulate the problems I'd been experiencing.

Apparently, ChatGPT was perfectly capable of helping me explain why ChatGPT wasn't working.

Armed with my very compelling complaint, I opened a support ticket.

Then came Emma.

Emma was OpenAI's support chatbot.

I never managed to connect with a live person.

Instead, I received variations of the same response, over and over.

Emma explained that ChatGPT was essentially functioning as designed.

And she wasn't budging.

Which was interesting, considering I was reporting that it wasn't functioning as it previously had.

I tried explaining the issue again.

Emma continued to explain how the system worked.

I wasn't asking how the system worked. I was explaining what had changed about my experience using it.

I found myself getting increasingly irritated.

With a chatbot.

How was I getting this frustrated with someone who wasn't even real?

At some point, I realized I was spending an extraordinary amount of time trying to get artificial intelligence to understand why artificial intelligence wasn't working.

Eventually, I stopped trying.

Later, I went back to check the status of my support ticket.

It had been closed.

Nobody had notified me. I hadn't received an explanation or any indication that my concerns had been addressed.

Apparently, the matter was resolved.

ChatGPT was working as designed.

I was still trying to figure out whose design we were talking about.

I had to laugh.

Not because the experience was particularly funny at the time, but because the absurdity of the entire situation was becoming impossible to ignore.

Here's what I keep coming back to.

We hear so much about the productivity gains associated with artificial intelligence.

And I've experienced them.

I've seen firsthand how valuable AI can be when it helps me think through complex problems, explore ideas, and accomplish work more efficiently.

But there's another side to that productivity equation.

What happens when the time you save using AI is consumed by the time you spend correcting it?

What happens when a tool you've integrated into your daily work becomes unpredictable?

And what happens when the very technology designed to make life easier becomes another thing you have to manage?

I don't have all the answers.

I'm still using ChatGPT.

I still believe it can be an extraordinary thought partner.

But this experience reminded me of something important.

I will always lead with my own voice, my own judgment, and my own experience.

If AI can enhance my thinking, strengthen my work, and help me run my business more effectively, wonderful.

But I cannot build a business that comes to a standstill when my AI thought partner forgets who I am.

And if it happens again?

I know exactly who to ask to help me write the complaint.