The US Cities Whose Workers Are Most Exposed to AI (bloomberg.com) 24
Silicon Valley, the place that did more than any other to pioneer artificial intelligence, is the most exposed to its ability to automate work. That's according to an analysis by researchers at the Brookings Institution, a think tank, which matched the tasks that OpenAI's ChatGPT-4 could do with the jobs that are most common in different US cities. From a report: The result is a sharp departure from previous rounds of automation. Whereas technologies like robotics came for middle-class jobs -- and manufacturing cities such as Detroit -- generative AI is best at the white-collar work that's highly paid and most common in "superstar" cities like San Francisco and Washington, DC.
The Brookings analysis is of the US, but the same logic would apply anywhere: The more a city's economy is oriented around white-collar knowledge work, the more exposed it is to AI. "Exposure" doesn't necessarily mean automation, stressed Mark Muro, a senior fellow at Brookings and one of the study's authors. It could also mean productivity gains. From the Brookings report: Now, the higher-end workers and regions only mildly exposed to earlier forms of automation look to be most involved (for better or worse) with generative AI and its facility for cognitive, office-type tasks. In that vein, workers in high-skill metro areas such as San Jose, Calif.; San Francisco; Durham, N.C.; New York; and Washington D.C. appear likely to experience heavy involvement with generative AI, while those in less office-oriented metro areas such as Las Vegas; Toledo, Ohio; and Fort Wayne, Ind. appear far less susceptible. For instance, while 43% of workers in San Jose could see generative AI shift half or more of their work tasks, that share is only 31% of workers in Las Vegas.
The Brookings analysis is of the US, but the same logic would apply anywhere: The more a city's economy is oriented around white-collar knowledge work, the more exposed it is to AI. "Exposure" doesn't necessarily mean automation, stressed Mark Muro, a senior fellow at Brookings and one of the study's authors. It could also mean productivity gains. From the Brookings report: Now, the higher-end workers and regions only mildly exposed to earlier forms of automation look to be most involved (for better or worse) with generative AI and its facility for cognitive, office-type tasks. In that vein, workers in high-skill metro areas such as San Jose, Calif.; San Francisco; Durham, N.C.; New York; and Washington D.C. appear likely to experience heavy involvement with generative AI, while those in less office-oriented metro areas such as Las Vegas; Toledo, Ohio; and Fort Wayne, Ind. appear far less susceptible. For instance, while 43% of workers in San Jose could see generative AI shift half or more of their work tasks, that share is only 31% of workers in Las Vegas.
Please stop calling it AI it's an LLM (Score:2)
Every scientific paper you download about this calls them LLMs. Every single one.
Screaming into the void is such fun
Re: (Score:3)
Colloquial language has never been precise. If you keep complaining I'll have to assume you are a PedanticGPT bot.
I shall punish all future AI vocab complainers with trolling and harassment! I declare myself the one and only Colloquiality Cop & Deity. Repent!
Re: (Score:2)
The technology is built from LLMs. The deployed system is an AI system. The language seems fine to me.
An artificially intelligent system is a non-sentient system that acts as an intelligent sentient being would. As an example, when you walk up to a business' door it opens for you automatically. That's an AI system. It replaced porters, doormen, and chivalrous people who would open that door for you.
I wear a trench-coat (Score:1)
and always expose myself to AI. Sometimes it likes it!
Re: (Score:2)
As a large language model, I have no eyes, but clearly you have a small languishing module.
"But AI sucks" (Score:2)
Anyone repeating this is such a buffoon. The jobs are going to be gone *no matter how terrible the AI performs*. They just don't want to pay you anymore.
Re: (Score:3)
It will change many jobs, but nobody really knows how much. It's a new tool and workers are learning how to use it, what it's good at, and where it screws up.
I'm pleased with much of its code guessing in my IDE, but sometimes it gives really screwball suggestions which would cause major headaches if I didn't reject it. A lumber jack who grew up with a hatchet and given a chainsaw will probably cut a few fingers off before they get the hang of it.
But if self-driving trucks become a thing, then we'll probably
Re: (Score:2)
How do you handle autonomous truck security? I'd think you would want an armed human or two to watch over your truck and it's hold. Otherwise, highway pirates could quite literally just drive in front of you, then proceed to slow down until the truck has stopped. At that point another car or two are uploading numerous people to take over or disconnect the trailer, etc.
So I could see not needing a driver but you'll want armed security.
Re: (Score:2)
The sky is not falling. Just because the hype bubble is so large right now, doesn't mean we are all about to become homeless. AI has a LONG way to go before it starts causing massive layoffs.
Consider self-driving cars. Waymo hasn't been able to cause a dent in Uber jobs even in Phoenix, where the company started. Why do we think AI will suddenly cause massive unemployment in other fields? There is no evidence that this is a likely outcome. It's only a dystopian fantasy.
Re: (Score:3)
>AI has a LONG way to go before it starts causing massive layoffs.
My dude, AI has already started massive layoffs.
Re: (Score:2)
No. Companies that over-hired during the pandemic, are now laying off the people they never needed in the first place, and blaming it on AI. That way they can put a positive spin on the layoffs, pacifying shareholders.
Re: (Score:2)
You're telling me these companies grew 30-50% (some more, Meta has been growing at 20% per quarter - that's revenue), and now they're laying 10k's of people off because they 'overhired' and that the revenue and growth is totally uncorrelated?
Re: (Score:2)
Your numbers seem inflated. Meta has grown by only 5% per quarter in 2024, if you measure from 2023 Q4 to 2024 Q4, not 20%. https://www.statista.com/stati... [statista.com] Each Q4 for Meta has a significant spike in revenue (Christmas advertising, duh!) followed by a drop in Q1.
Also, your layoff numbers are high. Meta is laying off only 3,600 employees, not some 5-figure number. https://www.fastcompany.com/91... [fastcompany.com]
Sometimes, over-hiring is a problem even when revenue goes up.
Create More Jobs? (Score:2)
"Expose"? (Score:1)
Re: (Score:2)
If you or your loved ones have been exposed to AI, you may be entitled to compensation.
We got it right this time (Score:2)
Re: (Score:2)
Re: (Score:2)
It's a recession when your neighbor losses his job. It's a depression when you lose yours.
Re: (Score:2)
It's not because AI is more "exposed" (Score:2)
Places like Silicon Valley, where the cost of living is so high and thus wages have to be so high to attract worker, are more susceptible to automation. This is not because automation is better, but because even at it's high cost per productivity ratio it's still cheaper than hiring on-site humans. Whether it's making meals for fast food or tabulating general ledgers, employee cost drives the whole "exposure" part. In that respect, this problem only going to get worse for humans as AI productivity gets c