...which, by the way, is how most automation turns out. I've worked in semiconductor manufacturing and software, and in both cases the automation that was at one point predicted to eliminate jobs, just made the people more productive, which made them more valuable, and instead of fewer jobs there were more, because more could be done by each person (so the range of applications that were economical to do was increased). So "it'll help, it won't replace you" is not pie-in-the-sky thinking, it's the most likely result, because it's what normally happens when automation improves.

It made them appear more productive. The machines just did the once basic work. Numerous studies have been conducted to quantify which technology made workers more productive. They all ended up with models too complex and variable to be practical.

Society just needs to accept a whole of our social truism is politically correct LARP to the drum beat of a financiers profit memes. What are a bunch of rich elders going to do if billions quit dancing like jesters for coin? Fight us from their far off estates?

800k sworn LEO in the US working for a paycheck, not fealty to a billionaire. The public has all the advantage but isn’t allowed to know it. Communal health is the future of work whether outsider financiers accept or not. We’ll just flood the inter tubes with AI media and crash the value of their copyright constraint system if they don’t like it. I have zero obligation to do-nothing daddy deep pockets who can point to a contract I wasn’t a party to. I don’t believe such people are worth coddling with my agency.

Did you generate this comment with GPT-3?

Ok in this case it did not make him more productive...

expand on communal health, sounds interesting

Artists which use AI and become more efficient will drive up supply of art and decrease its price. Additionally, some people who need art might just skip the artist entirely now, leading to less demand. This means some artists may no longer be able to make a living. As in, they're effectively replaced.

Or it will enable things that are currently impractical due to the amount of manual work involved - driving easy to produce stuff out, again.

We might be seeing AI-generated ilustrations instead of stock photos in web content. I for one would really prefer it.

Yes. I can see this being a killer (or killer app) for content libraries more than for replacing artisans.

It will help content producers avoid hiring to achieve good results.

I’m using AI to aid development of a game engine with a baked in model to replicate and help the user evolve the game world style, and rules, to their curiosity and tastes. Games will be even easier since familiar experiences are heavily constrained through painstaking play test. Game world generation parameters are finite. Even the art work we appreciate most is often constrained to a handful of style patterns; consider how flat web design was suddenly everywhere with only colors differing. Whole lot of the same goes into game asset production.

Gabe Newell called content creating AI an extinction level event for multimedia business. Of course politics as usual will be leveraged to create private monopoly of such technology unless open sourced.

Our best bet is such software is open source and thus aristocrats cannot lock it up behind artificial scarcity and copyright maximalism.

Extinction level event is precisely what I am looking at.

It’s funny to me that even here in HN people don’t get this at all. They fall behind simple truisms “automation only increases jobs”. It’s almost as if they’ve forgotten Trump got elected because vast swathes of America were deindustrialised and that led to a collapse to the quality of life in those regions.

On research that I’ve read, automation has definitely decreased the number of industrial jobs in America. Those people, some in large numbers, never found other jobs that were as well paying.

But like David Simon always says “American distopia doesn’t sell in America”.

People theoufht the same with advancements in agritech 150 years ago. If 30% of the population isn't out in the fields, they'll be unemployed forever. Turns out we invent new jobs all the time.

Whether they're worth taking Vs doing something more similar to your current job but for a bit less money is an individual's choice, of course.

The average height of Englishmen actually went down during the industrial revolution (presumably) due to malnutrition.

New histories of the period think of it as potentially as one of the few times we know where there was an enormous economic boom alongside a general collapse of quality of life.

Ir could be we are looking at something similar at the moment with the digital revolution.

Yes eventually we all retrain and the effect will be nil, but in the in-between time, a lot of people might suffer immensely.

I used that giant example a bit comically, perhaps. We're not talking about rapidly retraining a huge, illiterate, agriculture-focused workforce, with utterly game-changing technology.

We're talking about things that can be eased into, by white collar workers. And the numbers affected are nothing like the industrial revolution's either.

Im not sure for how long that will hold up if we end up automating all vehicle driving, image generation and robotising large parts of our workforce.

Driving is certainly a more significant workforce than image generation, but that seems to be a completely separate area of research.

Certainly. What I mean is - I didn’t expect image generative AI to get good this quickly. Before DallE I didn’t expect it to happen at all - I had no idea it was possible until it happened and I was forced to face it.

Driving seems like a tough nut to crack, but what if we suddenly start obliterating multiple areas of human employment one after the other? Ones we haven’t even thought of?

Or you could have the same turnaround and half the employees, guess which is cheaper.

It's even cheaper to not have any employees at all, if that's your goal.

but then you'd have no output to sell, and no business to be had.

They actually make an interesting point if we assume they meant no other employees than oneself. If I can make something entirely on my own with the help of AI and sell it, such as my own movie from Stable Diffusion text to video, then I'd make quite a lot of profit. It's no different than one person software companies, just extended to other forms of creation.

Why even comment at this point

Imagine you have a company. You can get your work force and double your gross or keep half your employees and leave your gross unchanged.

Which would you pick? Keep in mind you don't have to go through the headache of actually hiring people - all you need to do is not fire them.

Employees are my biggest expense by a wide margin. I could drop half my staff and increase my net overnight. I'm not going to double my customers nearly that fast so I'll be burning through payroll in the meantime... especially considering customer acquisition is going to get harder than employee acquisition over the next year.

Businesses die of indigestion more often then starvation.