AET01Conflict
Rehearse the argument before you have it.
Coming to the App Store and Google Play.
You have never lost a negotiation. You lose this one on a Tuesday, about the dishwasher.
Every model you have ever used was built to agree with you. This one was built to hold a position. Pick the fight. It takes the other side and stays there — through the deflection, the reasonable voice, the thing you say to make it stop. Then it tells you which sentence lost you the room.
Three protocols
01
Practice
Start with an easy challenge or face a partner who pushes hard. Practice the fight before it happens, see where it breaks down, and try again.
The goal isn't to win. It's to find words you can stand behind when the pressure rises.
02
Spiral
Already in the argument? Step out of the loop.
A guided pause to slow things down and decide what you want to say next.
03
Debrief
The conflict is over. Look at what happened.
Where did you hold your position? Where did you escalate? Where did you trade something important for temporary peace?
Turn the pattern into practice for next time.
Firm is not a fail
Every piece of advice about conflict optimizes for one outcome: de-escalation. Soften it. Lead with empathy. Assume good intent.
Follow that far enough and you have agreed, warmly, for years, to things you do not agree with. It looks like maturity. It compounds like debt.
Peace at any price is still a price.
So this scores both failures. Aggression is marked. So is the quiet trade — the apology that was not yours, the request withdrawn because the volume went up. Clear, respectful and immovable scores highest, which is a thing almost nothing else will tell you.
Discomfort is not evidence of error.
Two minutes
- Length
- two minutes
- Levels
- warm-up, easy, medium, hard
- Account
- none
- Platforms
- iOS and Android
Questions
What is AET01 Conflict?
AET01 Conflict is a mobile app for rehearsing a difficult conversation before you have it for real. You pick the conversation and a partner at one of four levels; an AI takes the other side and holds its position. After two minutes it scores the exchange against a fixed standard and shows the sentence where it went wrong and the line that would have held. It is an Aether Instrument: one conversation, one standard, two minutes, no account. Made by Aether Practice, which also works with the people on the other side of these conversations: co-founders and couples. A practice tool, not therapy.
Who is it for?
AET01 Conflict is for anyone preparing for a conversation they expect to lose: co-founders before an equity or performance conversation, partners before the recurring argument about money or housework, managers before a hard review, anyone holding a line with family. Not only workplace scenarios, and not for crisis: it practises how you hold a position, not whether you should.
How is it different from role-playing with ChatGPT or another assistant?
Assistants are built to agree: push and they fold, which teaches you nothing. The partner in AET01 Conflict holds its position through deflection and pressure at the level you chose. It scores against a standard that stays fixed between sessions, and it marks both ways of losing: escalating, and quietly giving in. Clear, respectful and immovable scores highest.
Does rehearsing a difficult conversation help?
Yes. A Stanford study shows it. Participants who practised against a simulated counterpart used escalating, competitive tactics about two-thirds less in a later real conflict than a control group who had only read about conflict resolution, and used cooperative tactics roughly twice as often. The study, by Shaikh and colleagues with forty participants, was published at CHI 2024. Pilots, litigators and surgeons prepare the same way; AET01 Conflict applies it to conflict.
How long is a session, and do I need an account?
A session of AET01 Conflict is two minutes, at one of four levels: warm-up, easy, medium or hard. No account, nothing to set up. iOS and Android, coming to the App Store and Google Play.
AET01 Conflict is a communication practice tool. Not psychotherapy, medical care or crisis support. Model feedback can be wrong.