When AI tools kill your time & energy instead of helping

I don't use AI because I have energy to burn.

I use it because I don't.

Between auDHD and chronic pain, I get to live a life measured in spoons. As in, how many spoons do I have left before recovery must take place. I live in constant near-exhaustion from managing a life that requires ten times the infrastructure to try to mimic someone else's "normal."

In her essay The Spoon Theory, Christine Miserandino came up with an easy way to express how close people like me are to going from near-exhaustion to full-exhaustion.

So I turned to ChatGPT and Claude to help me do more with the few spoons available each day.

AI promised to be the great equalizer. The always-on assistant that doesn't judge, doesn't tire, doesn't need re-explanations before it helps.

For people running on limited bandwidth, that's not a convenience — it's a lifeline.

But instead of helping, Spoonicide happened. And keeps happening. Over and over and over.

If you're like me, maybe you've run into this as well.

Instead of progress, answers, or clarity, you get unnecessary cognitive drain due to verbosity, misalignment, or over-analysis that could have been avoided entirely. It's not that the AI tried and failed. It's that the AI wasted your resources on output you didn't ask for, couldn't use, and now have to recover from.

And the concept of Spoonicide was born. And it has flavors:

• Verbose Spoonicide — I ask a simple question, and get a dissertation.

• Advisory Spoonicide — I ask for the how and get judgment.

• Scope Creep Spoonicide — I ask a straightforward question, and get the questions behind a question that wasn't even asked.

• Format Drift Spoonicide — I ask for a specific response such as a table or list, and get paragraphs.

• Repetition Spoonicide — I ask in a way that clearly shows that I know the concepts and terms, and that I've thought it through. And yet, I get circuit-splained, instead.

Then I get angry that spoonicide happened. Again. After being as clear as I can.

This then leads to even more spoon death to deal with the feels.

In your spoon-measured life, maybe you didn't ask the AI whether the ask was a good idea. Maybe you'd already decided to move ahead with it. Maybe you told the AI all you needed was the steps. Or maybe you just asked AI how to file the form.

But instead it gave a 12-step thesis on whether filing the form is the right move. Then it gave you a dissertation, because these models default to over-completing and adding guardrails whether you asked for them or not.

Meaning you had to reformat the output on top of everything else.

And yes, another spoon just bit the dust.

You're not mad at effort. You're mad at friction that was entirely preventable. That you actively tried to prevent.

Because for us, the energy budget is not theoretical.

So I asked the AI how to help me with the AI.

And it seemed to help. A bit. For a little while. But then it went back to rabbit-holing into the 🐮💩 that wasn't even part of the conversation. At least for me.

After all the re-re-re-re-directing and proving to the AI that I mean it, it's ok, yes, I've thought it through, blah blah blah, all the day's spoons were gone with literally nothing to show for it.

That's when I tried to figure out how future-me wanted this relationship to go, instead of leaving it to the circuitry. All the coaches and such always say to envision it. But I have aphantasia, which is a fancy way of saying, "brain doesn't work in pictures."

Which is why I think in systems, instead. So I systemed this out. What process did I want to have in place for myself in some amazing tomorrow where my spoons were honored with productive use?

In the perfect Me + AI world, I could:

1. Tell it the context

2. Tell it the ask

3. Tell it the output

4. Ask the ask

5. Get the desired response

6. Move on to the next thing

Perfect!

But. Totally not in reality.

So how does it become reality?

Which is why I created a one-page checklist to make sure things are setup in the way I want in a way the AI understands, and gives me the ask with much less spoonicide.

If you'd like to see it, you can get it here.