No matter how you slice it, for a good long while you will need highly skilled experts in those domains most affected by GPTs to be able to determine when the GPT gives a good answer or complete nonsense. Beware of the AI doomers who are convinced all software engineers will be unemployed in 5-10 years. But, also, be wary of AI polly-annas who see the rise of AI as purely wonderful. Neither of those extremes are going to be correct.
I'm just not qualified to try to read this. My CPA (1973) is just too old.
GPT? (OK, I looked that one up.) Very interesting. ALL tools require/reward peculiar skills. Ever watch an under-40 keyboard with their thumbs on a smart phone?
My guess for a while has been that AI will turn upper middle class jobs into middle class jobs as people who a high school education and a few months of training can do most skilled white collar work. Hence this reduces the wage polarisation in the labour market and reduces inequality.
Like humans, AI will be limited by its learning. But how much can AI understand? What data gives AI meaningful feedback to help it understand? Can it really develop knowledge, technical expertise and experience in itself? It may some day but it is a along way off. So enjoy the ride.
Nice piece but it may miss the real story: while inequality between different types of labour is bound to be reduced, inequality between capital and labour may sky-rocket (unless AI ends up being open-source and won’t give rise to winner-take-all effects). Thoughts welcome.
No matter how you slice it, for a good long while you will need highly skilled experts in those domains most affected by GPTs to be able to determine when the GPT gives a good answer or complete nonsense. Beware of the AI doomers who are convinced all software engineers will be unemployed in 5-10 years. But, also, be wary of AI polly-annas who see the rise of AI as purely wonderful. Neither of those extremes are going to be correct.
Made me think differently. thank you
Isn't there a historical analogy (steam power, computers, machines) that could be comparable to the rise of AI?
"That’s the demand side for abor"?
I'm just not qualified to try to read this. My CPA (1973) is just too old.
GPT? (OK, I looked that one up.) Very interesting. ALL tools require/reward peculiar skills. Ever watch an under-40 keyboard with their thumbs on a smart phone?
My guess for a while has been that AI will turn upper middle class jobs into middle class jobs as people who a high school education and a few months of training can do most skilled white collar work. Hence this reduces the wage polarisation in the labour market and reduces inequality.
Like humans, AI will be limited by its learning. But how much can AI understand? What data gives AI meaningful feedback to help it understand? Can it really develop knowledge, technical expertise and experience in itself? It may some day but it is a along way off. So enjoy the ride.
Nice piece but it may miss the real story: while inequality between different types of labour is bound to be reduced, inequality between capital and labour may sky-rocket (unless AI ends up being open-source and won’t give rise to winner-take-all effects). Thoughts welcome.