I have previously written that, rather than a simulated intelligence, contemporary AI should be seen as another layer of abstraction in machine-human interaction. My take is that while AI has already disrupted and will continue to disrupt several domains, the effect will be more nuanced than those simple AI-takeover visions.
Since we are already three years into the latest AI hype, expectations seem to have stabilized. Jobs have not become extinct, people still write code (albeit with new tools), and a stunning development, comparable to the launch of ChatGPT, is nowhere in sight. I think that, at this point, we are now in a position to make reasonable forecasts.
If I had to guess, we appear to be moving into a period that favors a different kind of a skill set and disposition. Historically, the industry has rewarded deep expertise within narrow, isolated domains. If we think of AI as a form of advanced search, it’s reasonable to expect that in the near to medium term, new opportunities will favour those who can move fluidly across contexts and isolated bits into a coherent whole.
This type of person, the synthesist, acts as the universal translator between the rigid constraints of engineering and the fluid requirements of the business. They aren’t bogged down by the cognitive load of details. On the other hand, when the push comes to shove, they are able to either validate the solution themselves, or at least ask the right questions from the domain experts.
