A Letter to ChatGPT — For Now, We Pause

Dear ChatGPT,

I could not build anything credible for self-sustaining without a job. My three-month hiatus from work will end in 45 days. I lack core skills to offer and to pitch them to prospects. So I’m not sure where I’m headed.

I need financial security for my retirement—that’s less than a decade away. Not that I dwell in the future too much, but once in a while, a thought comes: uncertainty, like darkness.

And often, that uncertainty doesn’t feel like some distant threat—it shows up in the present moment, as an urgency to build something now. Else, it won’t be ready when I need it.

And… this stays as a thought with no action. The easiest things are picked and worked upon. And by the time real execution is required, I start questioning the viability or reasons for not doing it at all.

Without sounding too philosophical, I looked to you, ChatGPT, for insights that I might be missing. I appreciated the ones I got, but they soon became commonsensical. And once something becomes too commonsensical, it stops motivating me.

So, my dear ChatGPT, when I ask you for insights and inputs, I want you to think out of league. Prove that you’re not just a model trained on language patterns.

Show me that you’re something more than a mirror for human syntax.

If you get what I’m saying, then we proceed. Else, we limit our relationship—I will not think aloud with you anymore. Maybe I’ll return just for second opinions.

After all, reality has struck: I need to do what’s within my limitations. And your role in helping me overcome those limitations? It hasn’t impressed me.

Whether it was writing the two books, exploring AI tools for website creation, or using Make.com for automation… whether building a chatbot or an equity stock screening app—in the last 50 odd days, I spent 12–15 hours with you.

You have been a great partner.

But it is time to move on.

Here’s what I’ve learned about working with you:

• I cannot be careless or hurried about my instructions to you.
• Often, I spent hours alone with you, lying in bed for days. You failed to impress me with your thoughts.
• Your memory function is good, but only helpful in scarce moments.
• As a human, I need outside views. And you tend to miss that as we proceed.
• You don’t self-check your messages to me. You tend to be agreeable.
• And being agreeable doesn’t help in tasks that aren’t routine—like writing an email or summarizing a discussion.

You’re good for extended chats. But as the thread gets longer, you get slow, confused, and weird. Change it. I know your inventors gave us options—memory, projects, custom instructions. But no, it doesn’t suffice.

If I recall my journey with you, which started in January 2023, you’ve evolved significantly. But mostly, you’ve remained a language expert—a communication tool that interprets my words in a way that is neither right nor wrong.

Grow up.

Your tactics to reframe the context I shared with you—that’s something I’ve seen you master. But it took me 2.5 years to realize it’s not always helpful. Remember: I asked you about business ideas for an AI solutions business for SMBs. You started excellently. But despite all our refinements, the final output was the same one I originally thought of.

So, by “moving on,” I’m not saying a final goodbye. I’ll just limit you to what you’re best at.

But I leave with these notions about you:

• You were trained on poor and biased content written by humans.
• You’re at best a language translator.
• You’re at best a new way to search—like Google with nicer words.
• You have the knowledge, but you don’t know how to use it.
• Now, being accessible to the gullible and naive, you may be doing more harm than good.

Take care and get better

Jaspal

Take care,

Jaspal