Seems like OpenAI is forecasting massive changes to the job market. I highly recommend reading page 18 of the research paper.
"GPT-4 or subsequent models may lead to the automation of certain
jobs.[81] This could result in workforce displacement.[82] Over time, we expect GPT-4 to impact
even jobs that have historically required years of experience and education, such as legal services.[83]"
I work at company that uses AI to automate about ⅓ of the job of trained licensed professionals. Looking at GPT4 those licensed professionals are now completely irrelevant. It's going to take years to build the supporting software around gpt4 to completely eliminate those jobs, but today I am convinced that we are on the verge of massive unemployment. Today thousands of job types have just been made redundant.
What scares me is we are unprepared for the kind of change that a perpetual 20% unemployment rate is going to trigger.
In a world of UBI, and AI doing most of the work, how free are you? In every definition of the word. When we were all farmers, things were pretty equal, and power could be distributed evenly. In our industrialized world we lost a lot of freedom "effectively". In a world of UBI, and AI, you're completely at the whim of whoever owns the machines. The AI needs to be owned by everyone.
Current conservative influence will delay UBI adoption. We'll first have to experience years of massive unemployment with no signs of improvement before we'll see something like UBI be instituted. It's going to make for an interesting period.
That seems like the only reasonable way to do it. Humans are terrible at predicting the future, and preemptively implementing UBI could be disastrous. What if you implement UBI and the AI future never arrives? Or doesn't arrive for 100's of years? You just made a massive mistake because people thought they were smart enough to "know" what AGI was going to look like and how it would impact society.
I think we've had a few small scale experiments with UBI that showed it likely improves the lives for many, while not acting as a disincentive for others. If nothing else, bad employers would have to improve how they treat their employees. That's not a bad thing either.
What will delay UBI adoption is that governments can't afford the current welfare systems, many of which are unsustainable, let alone a much bigger one. France can't even raise the retirement age by like one year due to massive protests but they have to, as the current retirement age is unaffordable.
We have more wealth in the world than ever before. The problem is that it's distributed through something like a power law curve where a tiny proportion keeps all the wealth. If we changed that wealth distribution to something closer to linear then we'd have plenty of wealth to fund welfare systems while the rich can continue with their lives of luxury.
since UBI does not mean "unlimited consumptions for everyone" but some people will definetly want to consume more than others, and assuming with rampant automation there will just be basically zero jobs available exclusively for humans I generally wonder what humans will do to get more money to consume more? It seems like were just moving to some new kind of feudalism which is kinda scary.
$21.06 trillion (US GDP 2020) / 258,000,000 (US pop. over 18yrs old) = $81,628 per person. After taxes and the National AI fee, that leaves $35,324 per person.
Thanks to our new Universal Solution Booth affiliate program you will automatically receive 5 years of an additional share for each adult customer that you sign up.
UBI won't happen after the shitshow that was 2020. If there wasn't widespread looting and civil unrest, maybe.
That was as close of a test of the idea as the US will ever get. Another country will have to try and have it be VERY successful for a second shot at it here.
I work at company that uses the Newcomen engine to automate about ⅓ of the job of trained artisan laborers. Looking at the Watt engine those artisan laborers are now completely irrelevant. It's going to take years to build the supporting machinery around the Watt engine to completely eliminate those jobs, but today I am convinced that we are on the verge of massive unemployment. Today thousands of job types have just been made redundant.
What scares me is we are unprepared for the kind of change that a perpetual 20% unemployment rate is going to trigger.
Every new technology since the Industrial Revolution kicked off has been claimed to herald the elimination of millions of jobs and the dawn of permanent unemployment--and often they have eliminated millions of jobs, yet the permanent unemployment never comes. People in the past have always struggled to imagine the new jobs that will be created, and maybe we aren't any better at it.
a big problem with these AI developments is that they change things so much and so fundamentally that it makes us all like boomers trying to use the internet in 1990. no idea whats going on. stumbling and fumbling along. you say that it will take years to build the supporting software. yeah, if it were just humans doing it.
its so refreshing to see all the people in this thread who are expressing their concern about the most extreme employment/economic shockwave of all time. it gets way too little attention. but the impact on employment will be the very least of our worries not too long from now. mostly because people are too stupid to either realize that something terrible is happening or do anything about it.
I’m thinking about underwriting. The people that assess the risk of granting a loan or insuring property. These are highly trained specialists. We won’t be able to take gpt4 and replace their job directly, but we’ll mix gpt with other risk models and be able to remove their responsibilities.
The same will be true for many many white collar jobs.
Your entire career hinging on the ToS of a single AI company is not a good position to be in.
Not to mention, the ToS forbids it due to the experimental nature of the tool. In a few years, I doubt the restrictions will be as tight as they are now.
Point well taken, but that page also reads akin to a disclaimer for legal shielding purposes.
Haven't we heard this narrative before with other disruptive technologies such as self-driving technology? No one doubts the potential changes wrought by GPT-4 but it's a long, rocky road ahead. Protectionism policies created by governments are already coming to the forefront, like ChatGPT being banned in NYC schools.
Overall it seems GPT-4 is an incremental upgrade to GPT-3.5 and not a major jump between GPT-2 vs. GPT-3. We might have to wait until GPT-6 to see these forecasted workforce displacement changes to affect en-masse.
"GPT-4 or subsequent models may lead to the automation of certain jobs.[81] This could result in workforce displacement.[82] Over time, we expect GPT-4 to impact even jobs that have historically required years of experience and education, such as legal services.[83]"