Mary Doria Russell, in her novel The Sparrow, has some interesting and potentially increasingly relevant ideas about professions and work relations of the future.
She speaks of a professional class, suggestively called "vultures", who, drawing on a mixture of cognitive and computer sciences supplemented by in-depth interviewing techniques, contribute significantly to the algorithmization and automation of what we currently call non-routine cognitive jobs (see esp. ch 8 & 10).
All she had to do was learn an astronomer’s job and then do it faster, cheaper and more accurately than he could do it himself. [ch. 8]
questioning him about his steps in learning a language, about tricks he used, habits he’d formed, methods he’d developed almost instinctively, as well as the more formal and academic techniques he used to analyze and understand a language, on the fly, in the field. [ch. 5]
It is not mentioned in the novel, but the official title of these specialists is Cognitive Task Analyst; the name they prefer to give themselves is Mind-Raiders, a slightly menacing wordplay on mind-reader.
The market for such services, besides companies trying to reduce overhead, would be cash-strapped research programs - SETI in the novel.
The subject, an astronomer, reluctantly cooperates with the Mind-Raider. Besides compliance, one of the reasons seems to be professional pride, love of one's work, and a desire to see it continued after the end of his career. The AI that will replace him is his legacy; as one of his friends tells him, "if a vulture does a poor job on you, you’re immortalized as mediocrity" [ch. 2].
Such motivation, quite powerful when present, cannot be relied upon by the organization. I would therefore expect contract clauses that secure the cooperation of mind-raided subjects to become commonplace. I don't expect to see strong organized and coordinated reaction to such clauses, considering that professions that are heavily regulated and have powerful professional bodies - law and medicine - have already been penetrated by AI, without significant opposition (at least none that I am aware of). I suspect that resistance is most likely to be individual; I wonder what the underlying ideologies would be…
Another user of the Mind-Raiding tools, but not necessarily of the services of the professional, would be a lone scientist, trudging alone in a dull, non-fancy field, whose vital importance for humanity he's the only one aware of. Another one is a lazy but brilliant scientist who develops an AI to help him fulfill his research requirements and / or publication quota - the ethical question here is who should get the credit for the deliverables? (Related to the latter case, see the extremely interesting debate started by a guy asking whether it is unethical to not tell his employer he has automated his job, or the case of a programmer who wrote scripts to secretly automate a lot of his job - and email his wife and make a latte)
Working alone, these two engage in what I call "self-reading"; since their expertise is most likely not in AI coding, it will require a self-programming system. The whole process might involve the subject using VR to interact with material objects through the sensors and appendages of the machine, so that the latter has immediate access to all the data it can process and use. VR would also allow the AI to create scenarios that, perhaps supplemented by brain scans, would help it clarify details and nuances of the cognitive processes of its subject / partner. Last, but not least, the subject, acting as a mentor, will describe his own thought processes and teach the AI problem-solving strategies via conversation.
Everything described so far implies awareness and consent from the subject. But we might also see involuntary / pirate Mind-Raiding, when the subject is put in a drug-induced coma or dream-like state, which mimics reality almost perfectly. The raider, in such a case, would be most likely reduced to observing the subject's actions in VR and his brain scans. Without the subject's full openness and cooperation, the algorithmization of his cognitive skills and processes will most likely be incomplete, but nevertheless good enough for some purposes.
Although I am aware of their dark side, mastering the tools of a Mind-Raider sounds like a lot of fun to me. Anyone else would like to develop a curriculum for training Mind-Raiders, or to participate in such a program? 😺