What it does
The thread follows Jev's launch on Hacker News.
Articles & threads
Developers discuss TypeSafe's claim that software needs an AI model for decisions, not another chatbot.
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
The thread follows Jev's launch on Hacker News.
Start by writing down a clear question and every allowed choice, such as deciding whether a customer wants a callback. A developer can then write code to connect your app to TypeSafe. The service quickly picks one allowed option instead of writing a chatbot reply.
Because the service only chooses from your approved answers, it cannot write out explanations or polite warnings. It will still pick an option even when a message is messy or confusing, so test it on tricky real-world examples before trusting it.