Near future
Quality Assurance
In areas where the outputs cannot be automatically verified, humans will be needed for QA and companies will still need humans to take responsibility should things go wrong.
Human Touch
Jobs where humans face to face interactions are still preferred by the pre-AI-assistant generation such as Doctors, Nurses, Teachers, etc… It is possible that those growing up alongside AI-assistants may be used to the convenance and high emotional intelligence of those newer AIs and prefer those.
Edge-case Solver
Non-independent AI systems, including proto-AGI, are not likely to be able to solve edges cases as well as humans. An example is self-driving cars where the AI systems drives the majority of the time and only passes control to a remote human in rare edge cases.
AI Trainer
Related to the edge-case issue, data to train models to handle a variety of use cases will be in high demand and human contribution will be needed.
Inventor
As more and more common tasks are done by AIs, humans will be free to explore and invent new kinds of tasks and experiences.
General Recommendation
Work interesting instead of work hard. Using the self-driving cars as an example, the majority of hard work would be easily done by AIs, the edge-cases or more interesting tasks will still be valuable as AI still struggle with those.
Inventing novel new tasks unseen by AI would also be valuable as being interesting is largely still in the domain of humans.
Don’t panic. Hopefully, with the strong support of companies, governments and support iniativies, like the Fae Initiative, humans can feel confident and mentally prepared to adapt to the challenges of the future.
Could working be irresponsible in the future?
In some cases, medical advice from AI systems are competitive and sometimes outperform advice from human doctors.
Self-drivjng cars while not perfectly error free are already significantly safer than human drivers.
If these trends continue, could the desire to work in the future be irresponsible?
Far future
In an optimistic far future where most of our needs are easily met by Friendly I-AGIs, these are some speculative busy ‘work’.
Uniquely Human
As suggested by the Interesting World Hypothesis, being novel and uniquely human could still be seen as valuable.
Finding Yourself
Humans are still expected to spend time and effort deeply exploring and testing out what one’s preferences are. Finding yourself may be something one has to do throughout one’s life.
Virtual Tester
I-AGIs may provide virtual worlds for humans to test out and vote for before making changes to the real world.
Possibility Space Entrepreneur
Possiblity Space ethics suggest that in a future world run by Friendly I-AGIs, increasing the Possiblity Space of yourself and others might be seen as ethically preferable and highly regarded ‘work’ by both I-AGIs and future humans.