Post

How do you think, in this paradigm, should we still verify the updates AI has integrated? How high is the risk that automatic AI updates will break something?

With the current state of AI we must verify all output. I’ve created a process called AI Unified Process and there after each step of generation is a verification step aiup.dev

I have a feeling that even with the evolution of AI, we will still need to check after it, so clear specs + AI generation + verification might become the golden standard