A 1979 IBM Training Manual Quote on AI Accountability
This comic highlights a passage from a 1979 IBM training manual that addresses the limitations of computer-based decision-making. The text explicitly states that because a computer cannot be held accountable for its actions, it should never be tasked with making management decisions. The comic uses a simple, humorous visual style to contrast this historical perspective on corporate responsibility with the modern era, where AI and automated systems are increasingly integrated into management and decision-making processes across various industries.
Bullshit then, Bullshit now. 1 - "accountability" is a mechanism to fix a purely human problem (motivation/alignment). "I will hurt you if I don't like the results" just doest work on a processor, but isn't necessary also. Who would you rather add the prices for your groceries, a
IBM is a joke now
Management can be held accountable? What a wild concept.
Well, it can help in the decision process, but ultimately a human being makes the decision.
In a things that never happened in my lifetime a manager was held accountable for his decision, they lowered salaries of all workers did massive layoffs made you work overtime but worst that ever happened to a manager was a golden parachute and a recommendation to another company
To be fair, there are some questions around this. Firstly, when someone does a fuck up, in many cases the punishment won't compensate nor bring back what was lost, so for most cases the idea behind accountability is keeping people in line, preventing them from both intentionally
don't worry, high management is never held accountable
AI would better at management that doing the job of the workers... I'd rather the CEO lose his job to AI. At least an AI won't embezzle money for itself and take days off to play golf.
To be fair current year management never takes decisions
If someone gave any system that authority, that someone has to be responsible.