> Companies cannot control their "AI" because their output is beyond the scale of their ability to QA. Right, now extend that thought to "replacing" programmer

> Companies cannot control their "AI" because their output is beyond the scale of their ability to QA. Right, now extend that thought to "replacing" programmers with AI. This is allegedly a scale at which we _can_ QA. Perhaps we reduce the job to humans QAing bot output as has been suggested by others. Now what happens when it fails QA and the bot doesn't come up with a satisfactory solution that meets the requirements? Perhaps the programmer has to... program? What about when the requirements change? Who performs the work for feasibility requests or exploratory project spike? Sounds like the programmer was not replaced by AI.