so if I read this correctly you foresee the split of "design" and "production" for SaaS software industry - where we similarly like ordering car today, will use "configuration calculator" to put up my desired vehicle and the sales role is now moved to assist me with that rather then persuate me that the current 1 model is the one for me :-) Interesting direction !!
"Compared to the other aspects that make software hardware I suspect the essential difficulty of forming a good and coherent opinion to embody as software will remain the domain of the best creative humans."
why would you think this would still hold if AI is really really good?
There is a general expectation that domains with fewer potential moves and faster or clearer scoring criteria will be more likely to see artifical super intelligence. Concise spaces like that can be explored in simulation, i.e. GoZero teaching itself the game. We see this in self-driving car development, in protein folding, etc.
"Good product ideas" seems to be at the opposite end of the spectrum. Its neither concise as a space to explore nor is it fast or easy to score an idea. Humans aren't neccessarily good at this, which is why product entrepeneurs are so rare and valuable. I am not forecasting that humans will get better at it per se, rather than I doubt AI gets to this before something like code writing where the space to explore is smaller and the validation happens at first test run.
so if I read this correctly you foresee the split of "design" and "production" for SaaS software industry - where we similarly like ordering car today, will use "configuration calculator" to put up my desired vehicle and the sales role is now moved to assist me with that rather then persuate me that the current 1 model is the one for me :-) Interesting direction !!
"Compared to the other aspects that make software hardware I suspect the essential difficulty of forming a good and coherent opinion to embody as software will remain the domain of the best creative humans."
why would you think this would still hold if AI is really really good?
There is a general expectation that domains with fewer potential moves and faster or clearer scoring criteria will be more likely to see artifical super intelligence. Concise spaces like that can be explored in simulation, i.e. GoZero teaching itself the game. We see this in self-driving car development, in protein folding, etc.
"Good product ideas" seems to be at the opposite end of the spectrum. Its neither concise as a space to explore nor is it fast or easy to score an idea. Humans aren't neccessarily good at this, which is why product entrepeneurs are so rare and valuable. I am not forecasting that humans will get better at it per se, rather than I doubt AI gets to this before something like code writing where the space to explore is smaller and the validation happens at first test run.