One problem with enterprise AI is that anything that makes sense in a 2012 business world sticks harder.
That’s one reason why “AI employees” as a term has such traction inside of large firms right now - they know what an employee is, they know the job title, they know the salary, and they can compare that salary to tokens.
Much easier when it’s a paradigm someone has known for decades, and they’re just AI-ing it.
Much harder when they have to also rethink the environment and system.