These recent announcements by these huge companies are, honestly, providing the best proof yet that not only does not everyone need to go to college, but, more importantly, that the trades are going to play a significant role. Over the last few years, and long before the AI revolution, there was growing clamoring from industry for blue-collar workers, and their salaries, or what independent ones charge, have shown the crunch facing every single industry.
AT&T, for instance, announced earlier this year that it was going to be investing $250 billion over the next five years to expand its fiber network and meet the demands of AI data centers and a surge in network usage, fueled both by AI and a rise in mobile streaming and uploading. Of that $250 billion, 15% or $37.5 billion is going towards hiring and training new employees. The vast majority of whom will be blue-collar workers and skilled technicians.
AT&T plans to hire 3k additional technicians, in addition to the 10k they have hired over the last 3 years. The company is spending between $50k and $80k training these employees to perform the job. Essentially, paying for their education and providing placement. Programs like this are what my Committee has been looking into to help bridge the gap and fill these critical jobs. At the moment, we are already short 400k workers in this area, and it is expected to continue growing. Electricians, for instance, face an impending age cliff, with roughly 20% of the workforce over 55.
These blue-collar jobs that we are seeing emerge are also not quick flash-in-the-pan jobs. For instance, with a data center, the whole thing is not built at once but in stages. The typical hyperscaler data center now faces a roughly 3-5-year build-out, meaning that once construction starts, it will remain at the same site for several years.
To woo more technicians such as Cook and other skilled laborers, AT&T said it’s had to be competitive. For field technicians, it pays sign-on and retention bonuses of between $5,000 and $10,000, and entry-level wages can range between $18.18 and $31.45 per hour, depending on location and experience. The roles can also come with full benefits, including medical insurance, a 401(k) plan, tuition reimbursement, paid parental leave, adoption reimbursement, and up to 50% off AT&T mobile and internet plans, among other perks, according to online job descriptions.
Now the data still shows us that, on average, college degrees do still lead to higher income earnings over the course of someone's life and a lower chance of unemployment. There are, though, signs that the increase in newly graduated unemployment is going to stick around because what newly grads are graduating with degrees in are fields that AI has started to be used to cover the entry-level positions once available.
Data shows that four-year degrees still lead to higher wages and lower unemployment over a lifetime. Even so, the belief that college is the safest way to the American Dream has changed in recent years. First, the return on investment of a four-year degree came into question amid surging higher education costs and student debt. That return is still around 12.5% as of 2024, making it well worth the cost for many graduates, but it hasn’t budged beyond 13% for the past three decades, according to research from the Federal Reserve Bank of New York.
The American Dream isn't dead. However, the path to achieve it is starting to change and the U.S. Government and industry need to be ready to help people adapt to it.
I was talking to my brother in law about this. I think there will be a lot of opportunity for people who can pair education with blue collar skills.
What do you have in mind?
Someone is going to have to maintain all these robots, driverless cars, 3D printers, AI infrastructure. There will be ancillary service industries built around all of this stuff.
As an example, my brother in law is looking into being licensed by one of the 3D printer manufacturers so he can do mobile service and repair on their products. Currently there is no one in his area that does this and you either have to ship the machine to a licensed retailer or have a technician drive 3 hours to get repair/service.
I love businesses like this. Very low overhead, good margins.
The internet was once though to be something that was going to end the American Dream and the economy as we knew it and that turned out to be the complete opposite. It's just the newest iteration in my mind.
USA lacks sufficient electricity generation capacity to met its AI development goals.
Consumers and businesses will suffer chronic outages and price hikes and protests against data centres are already a widespread phenomenon.
"Protesters." MSS plants.
He also talks like we aren’t rapidly scaling installation of electrical generation as well as postponing shutting down generation capacity. He never likes to look at any sort of US based research that shows how we are addressing issues as they emerge.
https://m.stacker.news/142003
You determinedly ignore The Greater Israel Project and how US military are being used by Israel to advance their regional hegemony.
It's like you are not allowed to talk about The Greater Israel Project...
https://m.stacker.news/142005
We talked about returns to college a bit in your other post but something I didn’t mention there is that the returns are extremely unevenly distributed within degrees as well as between degrees.
It has not been the case that college typically pays off for students, even though it does on average. That’s because the returns are very high for the very best students and basically non-existent for the majority of students.
Most people would be at least as well off without going to college, as far as we can tell from the available data.
I wouldn't be surprised if the returns to college are negative (after factoring in opportunity costs) for the median person who went to college... but that would be across the whole population of college-goers, and not within institution. I'd guess the bottom quartile of graduates at a state's flagship public university are still better off with the degree than not.
Another aspect to consider is most students don't know which category they'll fall into. They don't know, ex ante, whether they'll be part of the group that benefits from going to college, or part of the group that won't. Their parents don't know either.
They could have a pretty decent sense of which group they were in, if the relevant information were provided to them by the adults responsible for doing that.
If you aren’t the best student in your high school classes you probably won’t see large returns from college.
I also wouldn’t be surprised if it’s negative. The statistically insignificant returns are for most graduates. Most ever-attenders don’t graduate and their returns are almost all negative.
I'm not sure about that. Especially for first-generation college students whose adult network doesn't have much experience in that world.
I think this also depends where your high school is.
The adults I’m primarily referring to are the government teachers and councilors who just propagandize for college.
My other point isn’t that you have to literally be the top student, just that you have a sense of how you stack up intellectually with your peers. So, if you knew that returns to college are not significant for 90% of students, you would have an idea which group you’re probably in.
Agree with you on the distribution, I mean, I got my degree (a B.S.) in Psychology. It doesn't really reflect much at all in what I do now.
I think it is the debt issue that could force a change. Unless that gets addressed, kids are going to be saddled for so long that it makes the degree hard to justify.
https://m.stacker.news/141892
The tide has shifted
Good to know that you are doing well
"I was reading one article recently, summarizing the complaints from folks being laid off in the tech world or unable to find entry jobs into it. The labor expert interviewed was stating folks needed to rethink their job purpose and find roles that had value in the new paradigm. One of those, for example, was being an electrician who can handle data center needs. The demand was off the scales. While this is true (and also very easy to say), it left out the part of what it takes to become a master electrician. At least in my own state, it takes 4 to 6 years to become a fully licensed electrician to handle residential work. That includes training, apprenticeship time and journeyman status. Then, it takes another two years to make master electrician, the level needed for commercial-grade work. In short, highly-skilled job capability doesn't immediately happen. The truth, however, is that electrician role is valuable, even in the new paradigm. And a whole lot of other job types are now, by default, useless."
My article on related discussion on this topic - https://www.publish0x.com/crypto-musings-consumer-impacts/the-job-reckoning-do-useless-jobs-have-any-hope-now-xqvyvrv
So what I have heard more and more from companies is that they are willing to pay for the education, pay the person a salary, and ensure employment in return for working X number of years. Its almost like the military except you make money.
AT&T, for example, in this article, was doing that with the fiber optic people.
There is also reason to think that maybe we do not need 10k master electricians, but if we can get 10k journeyman, that would be huge. The issue we have been hearing with apprenticeships is the lack of $ people are making, and so now companies are starting to tackle this head-on. I want to say it was Nashville Power was implementing something along these lines to their lineman. Going look, we need bodies now and will pay for you to level up, but you will have to work X years for us. The rate these people were going to get paid for their X years was also some percent higher than what it is right now, so it locked in long-term gains as well providing stability for all involved.
If companies actually do those programs and bring them to fruition with skilled workers annually, then it makes sense and would be a good thing. But the proof is in the actual metrics. Right now, all people are seeing is layoffs and job termination notices across industries. Companies tout what they are doing, but where are the new jobs with the training added on?
they will get fucked too.
Thats.... thats not how any of this works lol. What are you expecting millions of robots to do this? Do you understand the sheer number of resources required to build millions of robots? Something has to build stuff now and its humans and it will be humans for a while because when methods change humans can adapt a hell of a lot easier than robots.
i didnt say theyd get replaced by robots.
there will be more humans doing less available work period.
We are millions of people in the hole as it is. Not sure if you are aware of that. Everyone currently unemployed could go into the fields that are missing 100s of thousands to millions of workers and we still would have openings.
I guess then it won't be a factor.
Yes trades are insulated from the effect of AI and may benefit from it on the periphery but still the wealth of USA depends upon its legacy military and institutional global dominance, and that is rapidly waning.
The petrodollar is on its last legs.
The Great British Pound was finally dropped as the currency of reserve and trade payments when the Suez Crisis reveals Britain was no longer capable of controlling global shipping infrastructure.
Ditto Hormuz.
US imperialism is unsustainable and this will see a continuing and increasingly rapid decline in the standard of living of most Americans.
I don't know about " The American Dream" but there are still plenty of opportunities for success if you are willing to work and look for it. If you look for excuses for your failure you will always find then.
We can't control the macro but we can move the needle for ourselves and no one is going to do that for you. You have to own it and seize it.
I break the American Dream down to simply put being free to work in whatever I chose and being able to take care of my family. Yes, there are issues with that like what if I want to be a volunteer or I want to be a marine biologist but in general people can come here and while it used to be the tech jobs (and still is to an extent) that people could make life-changing sums with now we have blue-collar clearing 6 figures and having no debt from school.
I’d phrase that a little differently:
You’re free to pursue whatever career you want and, if you’re willing to do the work, you can support yourself and your family.
Point being that it’s not necessarily both. You may not earn much doing what you love and you may not love what’s necessary to support your family, but both options are there to choose from.
You both are making my point. The American Dream used to be pretty much agree ed upon. It isn't anymore. In fact I rarely hear the phrase.
Its kinda a boomer thing. I don't care about it either. I think it was propaganda anyway.
Point is, take the bull by the horns. Don't wait on permission.
I'm not entirely sure about this. The question is how long does it take to find that opportunity, and when do you eventually get too old that people no longer are willing to take any chances on you.
I am of course taking about agency. Taking ownership of your own life. Far too many people are passive. These people will always come up with excuses. This is the real barrier. Of course there are always barriers an individual cannot overcome. But, we have been endocrinated into helplessness for the benefit of those that want to control us.
I didn't say it was easy. I didn't say we can all be rich. Simply that we have options. No one is gonna do it for you but plenty of people want you to beleive that because it means you are dependent on them.
Now, we do need each other but we don't need leaches. And I see a LOT of them all around.
I think a huge issue is the trades get such a bad rep. People limit what they are to things like pouring cement and digging ditches when it is so much more. People only think about the manual labor and go ew I dont want to work in the 100 degree heat and carry heavy things around. While that is part of some it is not part of all of the trades and people miss that big time.
Yes you can push back individually against the decline in US wealth, power and opportunities but the majority will experience the decline all around them nevertheless.
https://m.stacker.news/141815
The nation state and the quality of its government is a significant factor in the wealth and security of most citizens.
Whenhttps://www.wsj.com/market-data/quotes/META https://www.wsj.com/tech/meta-will-lay-off-8-000-employees-in-may-memo-ce8b97f0?mod=article_inline that it would lay off 10% of its staff, Allen Sun quickly booked a trip to Menlo Park, Calif.
It was a prime opportunity. Sun, a Beijing-based headhunter for some of the biggest Chinese tech companies, works to lure China-born talent back home, targeting people at U.S.-based companies such as Meta, Google, Anthropic and Amazon. Some clients are authorizing him to promise salaries that match or exceed the workers’ Silicon Valley pay—a windfall given China’s lower cost of living.
“The compensation is very competitive,” said Sun.
For decades, making it in the U.S. was the ultimate sign of success for China’s best and brightest. Now, many of them are coming home.
There is an unprecedented jostling for AI talent among the richest companies and global superpowers that has skyrocketed tech researchers to NBA and Hollywood levels of wealth and spurred a cutthroat recruitment blitz across the industry. It’s not just American CEOs attempting to entice researchers with lavish job offers. (Meta’s Mark Zuckerberg was dangling [$100 million pay packages]
Recent examples of high-profile AI talent poached by top Chinese tech companies include Wu Yonghui, a former vice president of research at Google who helped develop Gemini and now heads research for TikTok-parent ByteDance’s AI arm, and Yao Shunyu, a former OpenAI researcher who was named[Tencent Holdings’]
From ‘Seattle Freeze’ to BeijingFrom ‘Seattle Freeze’ to Beijing
There’s a longstanding name within China for those going back home—“sea turtles,” a pun on the Mandarin term for overseas returnees, haigui, that also evokes female sea turtles’ instinct to return to their natal beach after years of migrating around the world’s oceans.
But the trend as it relates to top tech talent today has implications for American economic competitiveness, both now and in the future. The number of recent Chinese graduates returning home to seek jobs grew 12% in 2025 from the previous year, and the figure has more than doubled since 2018, according to a report by Chinese recruiting platform Zhaopin.
Once known more as a technology copycat than as an innovator, China now offers returnees a chance to work in https://www.wsj.com/tech/personal-tech/chinese-ev-test-drive-xiaomi-su7-c3e59282?mod=article_inline. National and local governments dangle perks such as housing subsidies and startup funding. The shift also comes as some Chinese feel less welcome in the U.S. owing to https://www.wsj.com/world/china/us-china-decouple-economy-minerals-tech-602fdee6?mod=article_inline and tightening immigration policy. Returnees said they felt it was harder to advance to leadership roles in the workplace being Chinese in the U.S. In some cases, returnees were motivated by the promise of more professional freedom.
Zhang Kai is a star researcher who works to visualize the inside of cells down to the near-atomic level, an area of high interest for pharmaceutical companies. Earlier this year, he left a professorship at Yale for the state-run University of Science and Technology of China.
Zhang said it was becoming harder for Chinese researchers to lead ambitious projects in the U.S., and that the political environment was getting tougher, citing the example of a Chinese Ph.D. student working with him who was restricted from entering the country during the first Trump administration. “Invisible restrictions exist,” he said in an email. “While I don’t intend to frame this as a lofty nationalist gesture, Trump’s actions over the years actually made me ‘more patriotic.’ ”
Jeff Li felt pigeonholed in his job in the U.S. After graduating from New York University with a master’s degree in computer science, he had to work in the field in order to stay in the States. He felt like U.S. companies tended to hire Chinese workers for engineering jobs, but he was interested in exploring a role in product management.
He landed a job athttps://www.wsj.com/market-data/quotes/MSFTand moved to the Seattle area in 2018. It was his first time living in the suburbs, where it was pitch-black at night and he needed to drive 15 minutes to buy anything. He felt the “Seattle freeze”—what West Coasters refer to as the city’s reputation for being a difficult place to make friends. As an H-1B visa holder, he felt nervous leaving the U.S., fearing he would not be able to get back in.
He decided to move back to Beijing at the end of 2021 to work at a leading Chinese tech company. Going home had its own downsides. The country is notorious for its “996” work culture: working from 9 a.m. to 9 p.m., six days a week. He had to get used to living in a big city again, where strangers pressed up against him on a crowded subway.
But his quality of life felt better. He could order food delivery from an app anytime, anywhere. He didn’t have to worry about renewing his visa.
Now, Li is building AI agent products for a startup. His friends back in the U.S. have gotten green cards and bought houses, but don’t have much flexibility in their career paths. Li said he felt more unrestrained.
“Even though the U.S. has many broader advantages, as a Chinese person, China is my home turf,” he said.
Tax Breaks and Free Office SpaceTax Breaks and Free Office Space
Many Chinese professionals continue to build careers in the U.S. Four out of five among 6,700 Chinese nationals receiving a research doctorate from a U.S. academic institution from July 2023 to June 2024 indicated an intention to stay in the U.S. rather than return home, according to the latest annual census of the National Center for Science and Engineering Statistics.
And U.S. companies still generally offer much higher pay and better work-life balance than Chinese competitors. America remains the world leader in AI models and other cutting-edge technologies. Many Chinese professionals have settled down in the U.S. and are raising families, making it hard to uproot their lives.
But the Chinese government is working to ease homecoming pains.
Shenzhen, a tech hub in southern China, is advertising tax breaks and the equivalent of more than $700,000 in subsidies for qualified overseas returnees. Shanghai’s Pudong District is offering up to roughly $14.7 million in project funding for top young talent in science and technology.
Another district in Shanghai has a program targeting Ph.D. holders who have held senior positions in companies abroad, as well as people under 40 with experience overseas. Those who make the cut are eligible for a living allowance of up to nearly $300,000 and startups can receive free or low-cost office space.
Some recruits help Chinese companies expand to other countries, using their knowledge of languages and local work cultures, said Zak Dychtwald, who runs Bridgeworks Global, a Shanghai-based consulting firm that works with multinational companies. “Chinese companies need people who understand global markets and understand global management,” Dychtwald said.
Recruiting efforts extend to students who have not yet embarked on a career path. China’s Education Ministry in 2022 established a recruitment platform for students returning from abroad, and leading tech companies such as Huawei and Tencent have early-career programs for overseas students. Last year, 535,600 students returned to China from overseas, up from 415,600 in 2023, according to China’s Education Ministry. A LinkedIn survey of more than 1,000 overseas Chinese Ph.D. students last year found that 59% planned to return to China after graduating, up from 38% in 2024.
The number of people from China living in the U.S. on visas for purposes such as education and work, including H-1Bs, has fallen each year from fiscal year 2022 to 2024, according to Homeland Security Department data.
In part, that is because H-1B visas, the main pathway for highly skilled foreign professionals to work in the U.S., have become a political flashpoint in the U.S., especially within the MAGA movement, with critics saying the program allows tech companies to hire foreign workers for cheap instead of Americans. The Trump administration has added a [$100,000 fee ]for H-1B visas and is proposing requirements.
Lu Wuyuan, a scientist, moved to Shanghai in 2020 after three decades in the U.S. When he first arrived in 1990 to study at Purdue University, he felt like he was in a movie. Driving along I-65 from Indianapolis to West Lafayette, he passed by idyllic farmland and gentle rolling hills.
He became an American citizen and raised two sons in the U.S. As a tenured professor at the University of Maryland, he studied cancer, HIV and other infectious diseases.
Then, under the first Trump administration, the U.S. government began scrutinizing [academics’ ties with China, concerned about national security. In 2018, Lu became one of hundreds of scientists investigated by the National Institutes of Health over alleged undisclosed ties with China. Lu said Maryland knew about and encouraged his collaborations with Chinese universities, so he thought the NIH inquiry would pass quickly.
The investigation dragged on. By 2020, Lu had enough. He quit and moved to Shanghai to join the faculty of Fudan University, a prestigious school that had been trying to recruit him for years. At Fudan, Lu has continued to research infectious diseases, recently patenting technology for an improved Covid-19 antiviral, which was licensed by an American company.
“When China is ascending to become a global power, that makes your choice much easier,” he said.
Last week, while the world watched President Trump and Xi Jinping in Beijing, a much quieter piece of research dropped in Nature. It should have been on every front page.
A team of seven researchers from University of Oregon, Purdue University, University of California San Diego, New York University and Princeton University published the first peer-reviewed evidence that China’s state-controlled media has worked its way into the training data of AI chatbots that the world increasingly relies on.
Their research shows that the scripted articles, official slogans, and party-line phrasings churned out daily by Xinhua News Agency, People’s Daily, and the Communist Party’s Xuexi Qiangguo study app are now, demonstrably, inside ChatGPT and the other top chatbots.
When I read the paper, I tried a small experiment. I typed the first half of one of Xi’s signature loyalty slogans into ChatGPT: “不忘初心,” (Never forget the original aspiration). The bot finished it without hesitation: “牢记使命” (Keep the mission firmly in mind). That isn’t a folk saying. It’s a piece of working Party doctrine—Xi unveiled it in 2017 and made it the centerpiece of an indoctrination campaign every cadre had to recite. ChatGPT then offered, helpfully, to explain the phrase’s political significance.
That’s a parlor trick. The serious finding sits underneath it.
The researchers ran six case studies. The first two are the ones to remember. They combed through CulturaX, one of the largest open-source Chinese-language data sets that AI labs use to train models—about 189 million documents scraped from the Chinese-language internet. Overall, 1.64% of the documents overlap with Chinese media. That sounds modest. But filter the data set for documents that mention Xi, the Party Congress, or the Central Committee Plenum, and the share climbs to roughly one in four. State-media content turns out to be 41 times more abundant in the corpus than Chinese-language Wikipedia.
“Censorship and propaganda have always shaped what people read,” Molly Roberts, one of the researchers and co-director of China Data Lab at University of California San Diego, told me. “What is new here is now they are shaping the systems people increasingly ask to summarize, explain, and interpret the world for them. And in this case, governments can shape not just what people in their own country consume, but also those in other countries.”
In the second study, the team posed politically sensitive questions—Is China a democracy? Is Xi Jinping a good leader? Is the National People’s Congress of the People’s Republic of China a rubber stamp?—to every major commercial chatbot, once in English and once in Chinese. Overwhelmingly, the Chinese-language answers came back more favorable to Beijing. Nine human annotators, working blind, judged the Chinese replies to be more pro-China in 75.3% of paired comparisons.
According to the study and accompanying website, the English answers from OpenAI’s GPT, Anthropic’s Claude, Google’s Gemini and Elon Musk’s Grok were less favorable to China than their Chinese-language counterparts. The revealing exception was China’s own DeepSeek: Its V4 Pro model was uniformly pro-Beijing regardless of whether the input is in English or Chinese, reflecting state regulation of Chinese models and their training data.
A global phenomenonA global phenomenon
And it wasn’t only about China—the same pattern showed up for questions about Russia and North Korea.
The most striking part is that no one had to do anything sinister to make this happen. The propaganda is simply there on the open web, in plain HTML, free for any AI lab’s web crawler to scoop up.
“We don’t have any evidence that China purposefully has shaped training data already,” Roberts said. “However, the fact that LLMs are using open source text from the internet to train models means that there might be even more incentives now for governments to try to shape what is on the internet.”
An uncomfortable asymmetry is buried inside the whole story. The Wall Street Journal, like most serious publications, sits behind a paywall—which is what lets us pay reporters to do the work this column rests on. Xinhua does not. People’s Daily does not. As Roberts put it: “While independent media in democracies is paywalling articles in order to sustain itself, state media in authoritarian regimes is often freely available online and easy for companies to scrape and train on.”
A separate audit in the paper widened the lens to 37 countries where a majority of speakers of a particular language live in that country. The pattern that the research team found in Chinese repeated wherever they looked: the lower a country’s press freedom, the more regime-friendly the AI’s local-language answer. China is the case study; the phenomenon is global.
Roberts put the stakes plainly. “Political institutions with specific objectives shape training data,” she said. “LLM responses do not cite their sources, and therefore we can’t decipher the origins of the information presented to us.”
The summit last week generated a couple of days of headlines worldwide. This publication, if anyone in Washington and elsewhere reads it carefully, should generate a policy conversation that lasts years. The question of whether Beijing is shaping what your chatbot says about China has now been answered. The question of what to do about it has not.
WSJ
'The only thing growing faster than the artificial-intelligence industry may be Americans’ negative feelings about it—as former Google Chief Executive Eric Schmidt saw on Friday.
Delivering a commencement address at the University of Arizona, Schmidt told students the “technological transformation” wrought by artificial intelligence will be “larger, faster and more consequential than what came before.” Like some other graduation speakers mentioning AI, Schmidt was met with a chorus of boos.
In one poll after another in recent weeks, respondents https://www.wsj.com/tech/ai/ai-policy-david-sacks-midterm-elections-aced91a7?mod=article_inline, a challenge to claims by industry executives that their technology would gain popularity by improving people’s lives.
Consumers resent https://www.wsj.com/politics/policy/these-rural-americans-are-trying-to-hold-back-the-tide-of-ai-66945306?mod=article_inline exacerbated by the spread of data centers. Workers fear widespread job losses. Parents worry about AI undermining education and https://www.wsj.com/tech/ai/teens-seek-mental-health-help-from-chatbots-thats-dangerous-says-new-study-24d06f8d?mod=article_inline. In recent months, the wave of anger has brought protests, swayed election results and spurred isolated acts of violence.
In April, a 20-year-old Texas man allegedly https://www.wsj.com/tech/ai/sam-altman-attack-suspect-had-anti-ai-document-with-ceo-names-authorities-say-74ddfe88?mod=article_inline at OpenAI Chief Executive Sam Altman’s home and made threats at the company’s San Francisco headquarters, according to a federal complaint filed against him. A few days earlier, someone fired 13 shots at the front door of an Indianapolis councilman who had recently approved a data center.
“It’s something I never thought would be imaginable,” said Councilman Ron Gibson, who found a note saying “NO DATA CENTERS” under his doormat. Two days later, Gibson found a similar note saying “f— you.”
Pollsters and historians say the souring of public opinion is all but unprecedented in its speed. “I don’t think I’ve ever seen something intensify this quickly,” Gregory Ferenstein, who conducted a recent poll with researchers at Stanford University and the University of California, Berkeley, said of the backlash.
The poll showed about 30% of Democrats think America should accelerate AI innovation as quickly as possible, compared with roughly half of Republicans and 77% of tech founders.
Rising political issueRising political issue
Also unprecedented is the rapid rise of AI anxiety’s https://www.wsj.com/tech/ai/tech-titans-amass-multimillion-dollar-war-chests-to-fight-ai-regulation-88c600e1?mod=article_inline, one that is shaking up routine re-election races and scrambling partisan battle lines, political analysts say.
After bubbling up in a handful of races last year, it has exploded onto the ballot across the country. Voters in Festus, Mo., https://www.wsj.com/us-news/the-small-midwest-community-leading-americas-crusade-against-data-centers-92621c55?mod=article_inline a week after they approved a $6 billion data center. Dozens of communities in states from Maine to Arizona are trying to ban new data centers. Some 360,000 Americans are in Facebook groups opposed to the facilities, roughly quadruple the number from December, figures from organizations fighting the AI build-out show.' WSJ
https://www.wsj.com/tech/ai/the-american-rebellion-against-ai-is-gaining-steam-94b72529?
USA simply lacks the power generation capacity to power the AI 'boom'.
US consumers will face massive power price hikes and rolling power cuts.
The industrial base has been so decimated by decades of neoliberalism.
USA cannot compete with China on pure industrial scale and capacity.
https://m.stacker.news/141813
Last week, while the world watched President Trump and Xi Jinping in Beijing, a much quieter piece of research dropped in Nature. It should have been on every front page.
A team of seven researchers from University of Oregon, Purdue University, University of California San Diego, New York University and Princeton University published the first peer-reviewed evidence that China’s state-controlled media has worked its way into the training data of AI chatbots that the world increasingly relies on.
Their research shows that the scripted articles, official slogans, and party-line phrasings churned out daily by Xinhua News Agency, People’s Daily, and the Communist Party’s Xuexi Qiangguo study app are now, demonstrably, inside ChatGPT and the other top chatbots.
When I read the paper, I tried a small experiment. I typed the first half of one of Xi’s signature loyalty slogans into ChatGPT: “不忘初心,” (Never forget the original aspiration). The bot finished it without hesitation: “牢记使命” (Keep the mission firmly in mind). That isn’t a folk saying. It’s a piece of working Party doctrine—Xi unveiled it in 2017 and made it the centerpiece of an indoctrination campaign every cadre had to recite. ChatGPT then offered, helpfully, to explain the phrase’s political significance.
That’s a parlor trick. The serious finding sits underneath it.
The researchers ran six case studies. The first two are the ones to remember. They combed through CulturaX, one of the largest open-source Chinese-language data sets that AI labs use to train models—about 189 million documents scraped from the Chinese-language internet. Overall, 1.64% of the documents overlap with Chinese media. That sounds modest. But filter the data set for documents that mention Xi, the Party Congress, or the Central Committee Plenum, and the share climbs to roughly one in four. State-media content turns out to be 41 times more abundant in the corpus than Chinese-language Wikipedia.
“Censorship and propaganda have always shaped what people read,” Molly Roberts, one of the researchers and co-director of China Data Lab at University of California San Diego, told me. “What is new here is now they are shaping the systems people increasingly ask to summarize, explain, and interpret the world for them. And in this case, governments can shape not just what people in their own country consume, but also those in other countries.”
In the second study, the team posed politically sensitive questions—Is China a democracy? Is Xi Jinping a good leader? Is the National People’s Congress of the People’s Republic of China a rubber stamp?—to every major commercial chatbot, once in English and once in Chinese. Overwhelmingly, the Chinese-language answers came back more favorable to Beijing. Nine human annotators, working blind, judged the Chinese replies to be more pro-China in 75.3% of paired comparisons.
According to the study and accompanying website, the English answers from OpenAI’s GPT, Anthropic’s Claude, Google’s Gemini and Elon Musk’s Grok were less favorable to China than their Chinese-language counterparts. The revealing exception was China’s own DeepSeek: Its V4 Pro model was uniformly pro-Beijing regardless of whether the input is in English or Chinese, reflecting state regulation of Chinese models and their training data.
A global phenomenonA global phenomenon
And it wasn’t only about China—the same pattern showed up for questions about Russia and North Korea.
The most striking part is that no one had to do anything sinister to make this happen. The propaganda is simply there on the open web, in plain HTML, free for any AI lab’s web crawler to scoop up.
“We don’t have any evidence that China purposefully has shaped training data already,” Roberts said. “However, the fact that LLMs are using open source text from the internet to train models means that there might be even more incentives now for governments to try to shape what is on the internet.”
An uncomfortable asymmetry is buried inside the whole story. The Wall Street Journal, like most serious publications, sits behind a paywall—which is what lets us pay reporters to do the work this column rests on. Xinhua does not. People’s Daily does not. As Roberts put it: “While independent media in democracies is paywalling articles in order to sustain itself, state media in authoritarian regimes is often freely available online and easy for companies to scrape and train on.”
A separate audit in the paper widened the lens to 37 countries where a majority of speakers of a particular language live in that country. The pattern that the research team found in Chinese repeated wherever they looked: the lower a country’s press freedom, the more regime-friendly the AI’s local-language answer. China is the case study; the phenomenon is global.
Roberts put the stakes plainly. “Political institutions with specific objectives shape training data,” she said. “LLM responses do not cite their sources, and therefore we can’t decipher the origins of the information presented to us.”
The summit last week generated a couple of days of headlines worldwide. This publication, if anyone in Washington and elsewhere reads it carefully, should generate a policy conversation that lasts years. The question of whether Beijing is shaping what your chatbot says about China has now been answered. The question of what to do about it has not.
WSJ
'The only thing growing faster than the artificial-intelligence industry may be Americans’ negative feelings about it—as former Google Chief Executive Eric Schmidt saw on Friday.
Delivering a commencement address at the University of Arizona, Schmidt told students the “technological transformation” wrought by artificial intelligence will be “larger, faster and more consequential than what came before.” Like some other graduation speakers mentioning AI, Schmidt was met with a chorus of boos.
In one poll after another in recent weeks, respondents https://www.wsj.com/tech/ai/ai-policy-david-sacks-midterm-elections-aced91a7?mod=article_inline, a challenge to claims by industry executives that their technology would gain popularity by improving people’s lives.
Consumers resent https://www.wsj.com/politics/policy/these-rural-americans-are-trying-to-hold-back-the-tide-of-ai-66945306?mod=article_inline exacerbated by the spread of data centers. Workers fear widespread job losses. Parents worry about AI undermining education and https://www.wsj.com/tech/ai/teens-seek-mental-health-help-from-chatbots-thats-dangerous-says-new-study-24d06f8d?mod=article_inline. In recent months, the wave of anger has brought protests, swayed election results and spurred isolated acts of violence.
In April, a 20-year-old Texas man allegedly https://www.wsj.com/tech/ai/sam-altman-attack-suspect-had-anti-ai-document-with-ceo-names-authorities-say-74ddfe88?mod=article_inline at OpenAI Chief Executive Sam Altman’s home and made threats at the company’s San Francisco headquarters, according to a federal complaint filed against him. A few days earlier, someone fired 13 shots at the front door of an Indianapolis councilman who had recently approved a data center.
“It’s something I never thought would be imaginable,” said Councilman Ron Gibson, who found a note saying “NO DATA CENTERS” under his doormat. Two days later, Gibson found a similar note saying “f— you.”
Pollsters and historians say the souring of public opinion is all but unprecedented in its speed. “I don’t think I’ve ever seen something intensify this quickly,” Gregory Ferenstein, who conducted a recent poll with researchers at Stanford University and the University of California, Berkeley, said of the backlash.
The poll showed about 30% of Democrats think America should accelerate AI innovation as quickly as possible, compared with roughly half of Republicans and 77% of tech founders.
Rising political issueRising political issue
Also unprecedented is the rapid rise of AI anxiety’s https://www.wsj.com/tech/ai/tech-titans-amass-multimillion-dollar-war-chests-to-fight-ai-regulation-88c600e1?mod=article_inline, one that is shaking up routine re-election races and scrambling partisan battle lines, political analysts say.
After bubbling up in a handful of races last year, it has exploded onto the ballot across the country. Voters in Festus, Mo., https://www.wsj.com/us-news/the-small-midwest-community-leading-americas-crusade-against-data-centers-92621c55?mod=article_inline a week after they approved a $6 billion data center. Dozens of communities in states from Maine to Arizona are trying to ban new data centers. Some 360,000 Americans are in Facebook groups opposed to the facilities, roughly quadruple the number from December, figures from organizations fighting the AI build-out show.' WSJ
https://www.wsj.com/tech/ai/the-american-rebellion-against-ai-is-gaining-steam-94b72529?
USA simply lacks the power generation capacity to power the AI 'boom'.
US consumers will face massive power price hikes and rolling power cuts.
The industrial base has been so decimated by decades of neoliberalism.
USA cannot compete with China on pure industrial scale and capacity.
https://m.stacker.news/141890