LinkedIn on Wednesday, February 25, 2026
These notes are to stimulate discussion of the subject.
Without question, the ongoing development of Artificial Intelligence, including robotics, marks the onset of the most significant economic and social era in human history since the Industrial Revolution.
Unlike the Industrial Revolution which made people more productive, AI promises to replace a great proportion of today’s workers with intelligent machines – humanoid and other devices – the abilities of which are equivalent or superior to those of their human predecessors.
Everything is going to change as a result of this AI revolution. Gradually at first, but then at an increasingly accelerated pace. Describe it however you like, but the reality is that we are in the very early stages of creating a life form, in a manner of speaking, with which our biological species will, throughout our economy, be hard pressed to compete.
Not incidentally, extreme wealth will tend to accrue in favor of those individuals and their companies who develop and manufacture the AI, while the lower, ever growing, downward sloping element of the “K” economy suffers financially. This increasing concentration of wealth will continue until, in a more distant future, the AI become able to develop and manufacture themselves. What happens then is hard to imagine.
Initially, large, more successful firms who can afford AI will dominate economic growth. Smaller businesses will adapt as quickly as possible, leasing the technologies they need to stay competitive and that maybe too expensive for them to purchase outright.
I know how all this sounds, but the process I am describing is already underway.
How will people who might lose their jobs to AI make a living? The answer to this question depends upon the strategies adopted by our economy, organically, and by government. This transition – from an economy without AI to an economy with AI – is not something that a free market capitalist system will be able to accomplish on its own, the way it has adapted to computerization and to the Internet. Eventually, very substantial government regulation of the capabilities and deployment of AI devices will be required to protect the ability of a large majority of our national workforce to obtain employment at sufficient levels of personal income.
In the meantime, it makes sense that younger individuals in all strata of our society educate themselves to maximize their flexibility in the workplace. That they be committed to continuing on-the-job and third party re-education. And that they focus upon career paths which, to the extent possible, favor human capabilities and skills relative to which AI is currently inferior. In general terms, these would be jobs with significant elements of creativity, supervision, management, risk-taking and/or empathy.
Note that it is the destiny of artificial intelligence and robotics development, unlimited by significant government controls, to exceed the potential of its human creators in every respect. I wish I hadn’t said that, but I can’t imagine a set of circumstances in which it will not be true. Inspired and enlightened popular government will need to treat AI much in the same way we control nuclear and other potentially harmful technologies, including social media for example, limiting their development and deployment for the benefit of humankind. Will my children’s, grandchildren’s and subsequent generations be up to that task? I don’t know.
What about career paths that currently require substantial periods – perhaps years – of training given that AI may be able to perform relatively routine, early career tasks more cost-effectively than its human counterparts? Early career attorneys and doctors are examples of such professions, but there are many, many others that have the same structure.
Where, for example, will sufficient numbers of higher level attorneys who take depositions, negotiate and argue on behalf of their clients in court come from without enough human attorneys having been developed by the practical experience of being entry and mid-level lawyers? The simple answer for now is that law firm management will need to resist overusing AI to replace new, inexperienced attorneys at the risk of “de-humanizing” its entire profession or forcing its client base to depend increasingly upon less experienced, less qualified human counsel.
Where will judges come from to the extent that work now done by newly graduated attorneys is replaced by AI alternatives?
Clearly, career paths will need to be adjusted in favor of an emphasis on apprenticeship and on-the-job training at their employers’ expense – in addition to AI doing as much as it can for the employer – the costs of the extended training of humans to be passed along to consumers and/or government subsidized. Or perhaps, by mutual consent among economic players in certain sectors, affected firms will just agree not to use AI in the same way writers’ and other unions for “creatives” have done.
The problem we’re discussing is that AI not only replaces workers in less complicated jobs relative to which employees have minimum upward mobility. AI tends to replace the workforce from the bottom up, reducing the number humans who are educated by experience to eventually have the skills needed for higher level positions for which current generations of AI are unprepared.
-Les Cohen
