#NoFollowUpNecessary
A couple weeks back, I was arguing that when people come up with ideas, the satisfaction is in the telling, not in building. And I was making this statement generally for idea sharing. But then, I also mentioned that people share their "ChatGPT plan" with me now. Rather than sharing the idea, they share the business plan on how to achieve the idea, entirely generated by AI.
This resonated with several people who emailed me saying they have experienced the same thing. So someone they know has an idea, rather than risk the potential humiliation from being told their idea is bad, they share it with their favorite AI. Sycophancy being the default behavior, their idea is always validated. Even if not, the AI might suggest slight adjustment to the idea and make it valid.
And at the end of a conversation with ChatGPT, the LLM, trying to be helpful, will always ask something in the line of: "Do you want me to create a step by step plan to achieve this?" The answer is always yes. Please tell me how I can make millions off my unique idea, and give me the details... make no mistakes.
The plan that comes back is elaborate. You can even ask ChatGPT to expand on specific sections. Now, this plan is what ends up being shared. Every single time I receive those plans and read them, I notice something funny. When I ask a question about a section, my friends have no answers. They have to go back to the AI to get an answer. Why don't they have an answer? Because they are reading it for the first time with me.
Basically, because the plan is long and elaborate, who has time to read it? The satisfaction is in the format and complexity, not in the execution of the plan.
They had an idea, ChatGPT improved it, then it built a plan and solution for the problem. So their idea now has a solution, and the solution must be correct because AI came up with it. The problem is solved, we can file it in a cabinet. Executing it was never the issue.
I'm sure there will be a psychological term for this. A term for getting a psychological reward from watching AI come up with a plan of execution for your ideas. This isn't specific to OpenAI's ChatGPT, it's a catch-all for all generative AI in the current market.
Even when I'm doing research for a blog post, I'm often caught in the "Would you like me to expand further on this?" questions that can easily lure you into a rabbit hole. I guess AI providers are learning from social media. In social media, the goal isn't to socialize with friends and family anymore. Instead, they are trying to keep you engaged for as long as possible, to expose you to the maximum number of advertisements.
With AI, the goal isn't to impart you with knowledge. Instead it's to give you the satisfaction of appearing knowledgeable by keeping you engaged with an AI while they expose you to ads and spend your tokens.