Professional freelance translator here (English to two European languages).
You highly underestimate the amount of actual translating work we humans still have to do despite recent progress in automated translation. Automatically translated subtitles for films, series and lectures for major streaming platforms have to be entirely revised by humans.
Google is often one of my end clients and you'd be surprised to see how much they mistrust their own AI. They're extremely demanding on quality and have a human translator do a job (sometimes assisted with automated translation, sometimes not) then one if not two human editors review the stuff as per a very extensive and strict style guide.
Professional human translators are not disappearing anytime soon - and they're not just translating cheap user manuals.
This might be true for the time being, but isn't the ultimate objective here to eventually engineer an Artificial General Intelligence that is, by definition, capable of doing anything that humans do at least as well as humans can do them? If that's achieved, there's really nowhere to hide from automation.
Even if AGI were to happen anytime soon and assuming that it can do everything an intelligent human can, the world will start to crave for human experiences. Hypothetical example, in a world where robotic chefs can cook anything for you, the value of a human-chef cooked meal will exponentially increase. Along with technology, cultures and perspectives evolve too. I'm confident that we will get to AGI at some point, but it's a big 'IF' at the moment.
To a large extent that's already true, but the number of people currently employed as, e.g, boutique shoemakers or horse ranchers is pretty tiny, and there's really no entry-level employment in these sectors. I suppose the "everyone becomes a luxury craftsman" scenario wouldn't be the worst of all worlds and I don't know it's outright impossible, but the Wall-E/Idiocracy scenario also seems pretty plausible (look up "Sam Altman is not a blithering idiot"), and that's not even the worst-case scenario here.
Perhaps. But we don't know whether the objective of engineering AGI will be achieved and, if at all, when.
I'm only commenting on the current state of the industry and the way things are looking for the short to mid-term future, based on internal evidence. Some people think that translators have already largely disappeared from a lot of industries because of AI. Just saying that's not the case.
Good argument, Andrew. I like the O-ring metaphor and multiplicative examples.
I imagine the oft-foretold white-collar collapse will look like power laws taking shape. As the best workers get exponentially better using AI, they pull so far ahead as to render most of their peers ineffective and dismissed. However, as work per unit shrinks, so do unit prices, raising the overall quality of life for everyone.
The dismissed workers are not ineffective people. If anything, they represent the majority of labor today who have mismatched skills with jobs. In the hopeful and probable scenarios, they will find or create jobs that match their skills, leading to a more honest, efficient, and fulfilled labor force.
The speed of the first shift will determine the turbulence of the second.
This is an eloquent explanation of how artificial intelligence and mechanisation don’t immediately destroy jobs. Key word: immediately.
During the transition stage humans still add value by checking and tweaking. They also reassure customers who might be unnerved by a sudden wholesale switch, which might apply to your bank teller example.
But the human translation overseers and radiography checkers have an expiry date. They are training the machines to the point where the technology catches up. So yes, jobs will still be available, and may even grow. In the short term.
Just wanted to say that I came across your stuff by chance. The algorithm sent me here via Notes.
I enjoy your posts so much I went to buy the kindle. And at 2.99 it was a no brainer price.
The first line of the intro by Reid Hoffman had me.
"I strongly believe that great technology entrepreneurs aren’t just technology geeks, they’re also business geeks."
I'm probably not the first one to point out that the word "nerd" is probably closer to what you mean by a devotion to openness, speed, ownership, science, higher levels of personal autonomy, etc than geek.
But if I insist too much, that would make me a pedant and being pedantic is certainly not the nerd way (or TGW).
In any case, I really enjoy your book and hope that perhaps the TGW seeps into other fields beyond just tech and business. Fields such as humanities and politics. Then, I hope to get an update from you about TGW's progress in these fields as well.
Perhaps it's the geek's inner optimism to believe TGW has much to offer in these other fields as well.
Sorry I have to write my comment about your book in this post. I prefer this place than an Amazon review and to be precise (and we geeks want to be precise), I haven't finished your book yet.
Professional freelance translator here (English to two European languages).
You highly underestimate the amount of actual translating work we humans still have to do despite recent progress in automated translation. Automatically translated subtitles for films, series and lectures for major streaming platforms have to be entirely revised by humans.
Google is often one of my end clients and you'd be surprised to see how much they mistrust their own AI. They're extremely demanding on quality and have a human translator do a job (sometimes assisted with automated translation, sometimes not) then one if not two human editors review the stuff as per a very extensive and strict style guide.
Professional human translators are not disappearing anytime soon - and they're not just translating cheap user manuals.
This might be true for the time being, but isn't the ultimate objective here to eventually engineer an Artificial General Intelligence that is, by definition, capable of doing anything that humans do at least as well as humans can do them? If that's achieved, there's really nowhere to hide from automation.
Even if AGI were to happen anytime soon and assuming that it can do everything an intelligent human can, the world will start to crave for human experiences. Hypothetical example, in a world where robotic chefs can cook anything for you, the value of a human-chef cooked meal will exponentially increase. Along with technology, cultures and perspectives evolve too. I'm confident that we will get to AGI at some point, but it's a big 'IF' at the moment.
To a large extent that's already true, but the number of people currently employed as, e.g, boutique shoemakers or horse ranchers is pretty tiny, and there's really no entry-level employment in these sectors. I suppose the "everyone becomes a luxury craftsman" scenario wouldn't be the worst of all worlds and I don't know it's outright impossible, but the Wall-E/Idiocracy scenario also seems pretty plausible (look up "Sam Altman is not a blithering idiot"), and that's not even the worst-case scenario here.
Perhaps. But we don't know whether the objective of engineering AGI will be achieved and, if at all, when.
I'm only commenting on the current state of the industry and the way things are looking for the short to mid-term future, based on internal evidence. Some people think that translators have already largely disappeared from a lot of industries because of AI. Just saying that's not the case.
Good argument, Andrew. I like the O-ring metaphor and multiplicative examples.
I imagine the oft-foretold white-collar collapse will look like power laws taking shape. As the best workers get exponentially better using AI, they pull so far ahead as to render most of their peers ineffective and dismissed. However, as work per unit shrinks, so do unit prices, raising the overall quality of life for everyone.
The dismissed workers are not ineffective people. If anything, they represent the majority of labor today who have mismatched skills with jobs. In the hopeful and probable scenarios, they will find or create jobs that match their skills, leading to a more honest, efficient, and fulfilled labor force.
The speed of the first shift will determine the turbulence of the second.
I’m unconvinced. This is one example.
This is an eloquent explanation of how artificial intelligence and mechanisation don’t immediately destroy jobs. Key word: immediately.
During the transition stage humans still add value by checking and tweaking. They also reassure customers who might be unnerved by a sudden wholesale switch, which might apply to your bank teller example.
But the human translation overseers and radiography checkers have an expiry date. They are training the machines to the point where the technology catches up. So yes, jobs will still be available, and may even grow. In the short term.
I think that's an important factor as well.
Hi Andrew,
Just wanted to say that I came across your stuff by chance. The algorithm sent me here via Notes.
I enjoy your posts so much I went to buy the kindle. And at 2.99 it was a no brainer price.
The first line of the intro by Reid Hoffman had me.
"I strongly believe that great technology entrepreneurs aren’t just technology geeks, they’re also business geeks."
I'm probably not the first one to point out that the word "nerd" is probably closer to what you mean by a devotion to openness, speed, ownership, science, higher levels of personal autonomy, etc than geek.
But if I insist too much, that would make me a pedant and being pedantic is certainly not the nerd way (or TGW).
In any case, I really enjoy your book and hope that perhaps the TGW seeps into other fields beyond just tech and business. Fields such as humanities and politics. Then, I hope to get an update from you about TGW's progress in these fields as well.
Perhaps it's the geek's inner optimism to believe TGW has much to offer in these other fields as well.
Sorry I have to write my comment about your book in this post. I prefer this place than an Amazon review and to be precise (and we geeks want to be precise), I haven't finished your book yet.