A decision AI · almost ready
Should you stay or should you go?
Bring the career decision you can't stop thinking about. Forty-five minutes later you'll have a direction and a date already on it.
Roughly how the first minute goes
It goes live this week. Leave your email and you'll hear first.
You've spent two years thinking about this. It takes forty-five minutes to close.
The decision isn't easy. But thinking was never the thing that was going to finish it, and nobody has ever sat you down and made you answer the questions in order.
I want in earlyOpens this week. First access goes to the list.
Which one of these is you?
I keep waiting for the next raise to make the decision for me.
I've been going back and forth on the same decision for over a year.
I stayed too long at my last job. I'm not doing that again.
I don't know if my side thing is a business or an expensive hobby.
The decision I was stuck on in January is the exact same one I'm stuck on now.
I've got a job I don't hate and a side thing I can't grow.
I know what I want to do. I'm just waiting for my partner to be okay with it.
I've made the spreadsheet. Twice.
More thinking isn't what's missing. You've thought about this ten thousand times. You don't have an information problem, you have a process problem, and no amount of new input has ever fixed one of those.
By the end of the session you'll have
…and the document that holds all of it.
↓ Here's exactly what happens, phase by phase ↓
If you can't even name the problem yet
Most people are one ring off. They're about to make a career-sized change to fix a job-sized problem, and no amount of deliberating fixes an aim that's wrong.
Frame it
What the decision actually is, how long it's really been open, and what you've done about it rather than thought about. Most people undercount the years by about half, and it pushes once.Diagnose the discontent
One question does most of the work: what would have to change for you to happily stay another two years? The answer tells you whether you hate the work or you hate this company, and those are completely different decisions.Widen, then narrow
You arrived with two options. There's almost always a third, and it's the one nobody has priced: something real changes and you stay.
If you know the problem and still can't move
This is where the session stops being a conversation and starts being a machine. Phase five produces a number, and that number decides phase seven.
Price both paths
Leaving costs something. So does staying, and almost nobody adds that column up. Money, time, energy, opportunity, and who you'd have to stop being.Score your agency, then your confidence
Five dials, weighted for the season you're in, because a 24-year-old and a 41-year-old with a kid shouldn't be turning the same ones. Then the question the whole thing hangs on: how sure are you, honestly, that the alternative delivers what you want?Classify the crossing
Door, bridge, or cliff. Then the harder half: how have you been treating it? The gap between those two answers is usually the entire problem.Build the bridge
Whatever you were least sure about in phase five is the thing to go find out. That link is the mechanism. It turns your experiment from advice into the obvious next move.The map, and the date
All of it, filled in with your words, in one document you can copy. Plus a first step small enough to do in the next two minutes.
Phase five deciding phase seven
Some of you will finish this and decide to stay. That's a real outcome and it'll tell you when it thinks it's yours. The point isn't to get you out. It's to make staying a choice instead of a default.
You finish holding this
A chat evaporates when you close the tab. This is the part that survives Monday, filled in with your answers, ready to paste anywhere you'll actually look at it.
THE DECISION MAP 1 · THE REAL PROBLEM It isn't the work. It's the company. If I had a manager who trusted me, I'd seriously consider staying. 2 · THE PRICE Stay 24 more months: I gain stability I give up my thirties Leave: I gain time I give up $40k and the title The price I'm more willing to pay: the money 3 · WHAT MATTERS, AND HOW SURE I AM What matters most right now: autonomy and time Current path 4/10 · Alternative 8/10 My confidence the alternative delivers it: 40% 4 · MY CROSSING This is a BRIDGE So I should TEST 5 · MY NEXT PLANK Biggest untested assumption: that I can bill $6k a month freelance Cheapest way to test it: land one paid project at that rate I'll have done it by: 14 September THE TWO-MINUTE MOVE: message the three people who've already asked THE STRETCH: tell my manager I'm going to four days
An example, not a template. Every line comes back in your words, and it never leaves one blank.
What you're actually getting
The whole session, written as instructions the AI follows. It asks permission before it pushes, and then it pushes. Roughly fifteen to twenty exchanges.
A one-click ChatGPT link and a Claude file. Same protocol, whichever one you already have open. Works on the free tier of either.
Rings, pricing both paths, the agency audit, door/bridge/cliff, and the plank. Running underneath the questions, so you feel them instead of studying them.
Annie Duke, Daniel Gilbert, Daniel Pink, Barry Schwartz, and 142 of my own interviews. Cited by name when it matters, never used as an opener.
The document above, filled in with your answers, with a date on it. This is the thing you keep.
One payment, no subscription. Use it again for the next decision, or hand it to a friend who's been circling one of their own.
What this costs instead: a career coach runs $150 to $300 an hour and you'd want three sessions. A decent therapist won't touch the logistics of it. Another year of deciding costs you a year.
Yours for $49
Why the usual advice hasn't worked
What you've been told
- Get clear first, then act
- Follow your passion
- Wait until you're ready
- Take the personality test
- Fix the résumé, find the next job
- Trust your gut
What actually closes it
- Clarity comes out of the experiment, not before it
- You can love the work and refuse the life
- Ready never arrives. A date does
- Your imagination is a bad witness. Go and find out
- Close the decision first or you'll take the wrong job
- Your gut knows the pattern, not the facts
But couldn't you just ask ChatGPT?
You could, and you probably have. Here's what happens differently.
It's built to agree with you
Ask a chatbot whether you should quit and it'll help you feel good about whatever you already said. This one asks permission to push, and then it pushes.
It interrogates instead of explaining
You've read the articles. What you haven't had is somebody who keeps going after the answer you always give, because that answer is the reason you're still here.
It ends with a document and a date
Most AI conversations feel great and change nothing. This one closes with an artifact, a dated experiment, and a line asking you to come back and report.
Where a normal chatbot would have agreed with you
And yes, it's a prompt. A long, specific, heavily tested one. There's a free, lighter version of it at kiaorion.com/otto with every word public, so you can hear the voice before you spend anything.
What $49 buys is the full eight-phase version, the frameworks underneath it, and the map at the end. Everyone else in this business would have hidden the free one.
Every decision is one of three things
Cheap, and you can walk back through it.
Take the recruiter call. Publish the thing. Ask to go remote.
Open it.
A real change, but you can test it in pieces.
Pivot careers. Go freelance. Move countries. Start the business.
Cross it a plank at a time.
Expensive, and very hard to undo.
Quit with no runway. Sell the company. Spend $150k on the degree.
Stress-test it first.
The expensive mistake is treating a door like a cliff, and most people have been doing it for about a year. The reverse is rarer, worse, and nobody sees it coming.
Who it's for, and who it isn't
For you if: you're employed, you've got something on the side or a place you keep almost moving to, and you've been going back and forth for over a year.
Not for you if it's been under three months. You haven't been stuck, you've been thinking, and that usually sorts itself out.
Not if the decision belongs to someone else. It can't help you decide what your partner should do about their job.
And not if something just happened. If you're in the middle of a redundancy or a breakup, this is the wrong tool, and it'll say so and stop rather than run the protocol on you.
Why me
I'm obsessed with how people make decisions. Not in a productivity way. In a “how does somebody spend nine years on something they quietly stopped wanting” way.
I spent about ten years making music that never broke. I don't regret going after it. What I regret is going after it in a way that built nothing I could use afterwards. It ended and I was left holding almost no skill that transferred. If I could go back I wouldn't tell myself to want it less. I'd tell myself to chase the exact same thing in a way that paid me something either way.
I've also spent about ten years living abroad, and roughly half of that deciding whether to go. Since then I've talked a lot of people through their own version of it. Whether to move. Whether to keep the job. How to keep the job and move anyway.
This is that conversation, built into something you can run at 11pm on a Wednesday without booking anybody.
I don't want you to build my version of a good life. I want you to get better at choosing yours.
Kia Orion
If it doesn't work
Thirty days, your money back, no questions. Email me inside a month and I'll refund you. You won't have to explain yourself and I won't try to talk you out of it.
Run the whole thing first, though. The session only does its job if you get to the map at the end.
Questions
Do I need a paid ChatGPT or Claude account?
No. It runs on the free tier of either. A paid account handles a long session more smoothly and remembers more of it, so if you already pay for one, use that one.
How long does it take?
Thirty to forty-five minutes, roughly fifteen to twenty exchanges. One sitting is better because it builds on your earlier answers, but you can stop and come back and it picks up where you were.
Is this a subscription?
No. One payment, and the files are yours. Run it again for the next decision, or for a friend's, as many times as you want.
What if my decision isn't about work?
It's built for career decisions. Stay or go, this job or the other one, the side thing or the salary. Bring it something else and it'll tell you straight away that you're outside what it was made for, then offer to carry on anyway with that caveat. It won't quietly run the wrong process on you and let you find out at the end.
Will it tell me what to do?
It'll tell you what it thinks and why, and it'll tell you when it thinks you should stay. What it won't do is hand you a pros and cons list, because that's what you've been doing alone for two years and it hasn't worked.
What if it tells me to stay?
Then you've got two years back and a reason. I've got no stake in you leaving, which is more than most people selling you something in this category can say.
Is my session private?
It runs in your own ChatGPT or Claude account. I never see it. If you want me to, you have to send it to me.
The question isn't whether $49 is a lot.
It's whether you want to spend another two years not answering this.
Get first accessOpens this week. First access goes to the list.
Want it the day it's up?
It's days away. Leave your email and I'll tell you the moment it's live. That's the only thing I'll use it for.