I agree that more generalists can succeed. That said, I think your argument assumes that the answers you get from ai are *correct* and *easy to get*.
In many domains, the correct answer isn't decided or readily available. For example, what's the best way to structure content for a marketing page? There are numerous theories and studies proposing wildly different strategies. In these areas, personal experience and even hunches can carry weight.
And getting at an answer, even with the best available models, is often not as massive of a leap from old-school googling as I might expect. I end up in long back and forth conversations with the llm and clicking out to references to read further. Is it easier and better? Probably. But it's definitely not as easy as "Ask the question. Get the answer."
The advice I've heard, even before ai, is to be both. Get a general understanding of a space and also go deep within a narrow piece of it.
I just finished reading "Range: Why Generalists Triumph in a Specialized World" by David Epstein. If you haven't read it, I think you'll find it fascinating.
I think it's important that you characterized "generalist" as someone with 4-6 domains of working knowledge. I think that's right. I def see people combining a few different areas of working expertise and knowing when and how to apply each as top skills in the A.I. age. I don't think that just being generally well-read will get the job done. Expertise still matters. In fact, combining several areas of expertise (4-6) could become a new area of speciality. Specialization and division of labor happen one after the other, over and over.
“When knowledge has depreciated to near zero, what is valuable?” My first thought was ‘attention’. What do we choose to give our attention?
But then I shifted. I believe relational connections will be valuable. The economy still operates in circles of relational trust.
Yes, persuasion, trust, and wisdom are appreciating rapidly.
I agree that more generalists can succeed. That said, I think your argument assumes that the answers you get from ai are *correct* and *easy to get*.
In many domains, the correct answer isn't decided or readily available. For example, what's the best way to structure content for a marketing page? There are numerous theories and studies proposing wildly different strategies. In these areas, personal experience and even hunches can carry weight.
And getting at an answer, even with the best available models, is often not as massive of a leap from old-school googling as I might expect. I end up in long back and forth conversations with the llm and clicking out to references to read further. Is it easier and better? Probably. But it's definitely not as easy as "Ask the question. Get the answer."
The advice I've heard, even before ai, is to be both. Get a general understanding of a space and also go deep within a narrow piece of it.
I just finished reading "Range: Why Generalists Triumph in a Specialized World" by David Epstein. If you haven't read it, I think you'll find it fascinating.
I think it's important that you characterized "generalist" as someone with 4-6 domains of working knowledge. I think that's right. I def see people combining a few different areas of working expertise and knowing when and how to apply each as top skills in the A.I. age. I don't think that just being generally well-read will get the job done. Expertise still matters. In fact, combining several areas of expertise (4-6) could become a new area of speciality. Specialization and division of labor happen one after the other, over and over.