You’ve done it. You’ve stared down a career change, a breakup, a business pivot — and typed “help me decide” into ChatGPT. And what did you get back? A bulleted list of generic pros and cons that you could have written yourself, except worse, because at least your own list would have had some soul.
Here’s the uncomfortable truth nobody tells you about AI advice: the reason it feels hollow isn’t because the model isn’t smart enough. It’s because you outsourced the only part that mattered — the thinking.
A developer who got frustrated with this exact loop built a tool called Reloadium Decisions. It’s deceptively simple. You don’t just ask for advice. You’re forced to lay out your assumptions upfront. You run a premortem — imagining the decision has already failed and working backward to figure out why. And you think across three time horizons: what happens in a week, a year, five years.
Notice what’s missing? The AI’s opinion.
The tool isn’t designed to give you answers. It’s designed to make you earn your own conclusion.
Here’s where it gets interesting. That premortem feature? It’s not some innovation born in a Silicon Valley hackathon. It’s Gary Klein’s cognitive science research from the 1990s, which itself draws on Stoic negative visualization — the ancient practice of pre-feeling failure so you can navigate around it. The three time horizons? That’s straight out of strategic foresight frameworks that militaries and intelligence agencies have used for decades.
What we call “AI decision-making” is really just ancient wisdom wrapped in a chatbot interface.
And that’s exactly the point. The tool works not because the LLM is brilliant but because it forces you to do what you were avoiding: naming your assumptions, confronting worst-case scenarios, and stretching your timeline beyond this week’s panic.
The relief you feel when using it isn’t from the AI solving your problem. It’s from structure replacing chaos. Paralysis isn’t a lack of information. It’s a lack of framework.
This is why most people misuse AI. They treat it like an oracle — tell me what to do — when the real value is as a scaffolding device. The model doesn’t know your life. It doesn’t know that you’re secretly hoping the decision goes one way because you’re terrified of the other. It can’t surface the assumption you’re refusing to name.
But a tool that says list your assumptions before you get any analysis? That forces the issue.
The best AI tools don’t replace your judgment. They make it impossible to hide from it.
So the next time you’re stuck on a hard decision, go ahead and use AI. But don’t ask it for the answer. Ask it to make you think harder. The difference is everything.
FAQ
Q: Isn't this just a fancy journaling prompt with a chatbot attached?
A: Yes, and that's the feature, not the bug. The structure — assumptions, premortem, time horizons — is what does the heavy lifting. The LLM is just a wrapper. The thinking is yours.
Q: So should I use this tool or just think harder on my own?
A: Use it. Raw thinking rarely forces you to confront what you'd rather ignore. A structured framework makes you name your assumptions and imagine failure before you commit. That's the difference between reflection and decision-making.
Q: Isn't the whole point of AI to automate the hard work?
A: For tasks, yes. For decisions, no. Automating your thinking is how you end up with confident-sounding advice that doesn't fit your life. The best AI decision tools don't replace judgment — they make it impossible to skip.