Essay by Eric Worrall
The USA was on track to secure its position as the AI center of the world. Then AI-hating Democrats won the midterms.
A Big Loser at the Ballot Box: New Data Centers
By Eliot Brown
ReporterIt’s a rough morning for data centers on the ballot.
Gubernatorial candidates pledging statewide data-center pauses won in Texas, Georgia and Michigan, while voters overwhelmingly passed moratoriums on data center construction in an array of local city and town ballot initiatives around the country.
Data-center stocks slid premarket; keep reading for more details on the reaction.
…
The stakes are high in Texas, the country’s epicenter of data-center construction, as Democrat Gina Hinojosa defeated Republican Greg Abbott with 52% of the vote. Hinojosa made opposition to data centers a key part of her platform, pledging a moratorium on new construction, pending new legislation that gives local communities greater say in future construction.
Ohio, another hub of data center construction, elected as governor Democrat Amy Acton, who pledged a conditional moratorium on data centers, saying that new servers needed to meet labor and energy requirements.
…
Read more: https://www.wsj.com/livecoverage/stock-market-today-dow-sp-500-nasdaq-11-04-2026/card/a-big-loser-at-the-ballot-box-new-data-centers-lUJTT0Zz4iGxfrmRUSN4
AI does not have to be located in the USA to shake up US jobs. Business owners will just plug into Chinese data centers instead.
Lots of businesses were already opting for the Chinese AI option before the midterms, because of China’s low-cost energy advantage. AI cost is dominated by energy cost. But President Trump’s enthusiasm for AI and for affordable US energy was keeping the USA in the game.
Chinese AI models surge in global popularity — and Washington is worried
PUBLISHED SAT, SEP 26 2026 1:00 AM EDT
Kai Nicol-Schwarz @IN/KAINS…
AI was a major focus as U.S. President Donald Trump and Chinese President Xi Jinping met this week.
The most advanced U.S. models lead most benchmarks, but Chinese companies like DeepSeek, Z.ai and Alibaba have released new models with major performance gains in tasks such as coding.
According to usage data shared with CNBC, the global adoption of Chinese models by companies has increased substantially in 2026.
Chinese AI models have gone from a relatively small share of usage to a majority on two major developer platforms that provide gateways for companies to access models from different providers. On OpenRouter, they accounted for 57%-67% of tokens used in the week of Sept. 14, up from 6%-13% in February. On Vercel, their share rose to 55% in August from 11% in January.
…
Read more: https://www.cnbc.com/2026/09/26/china-ai-global-adoption.html
Those figures measure which models businesses are choosing, not where the servers sit. But the economics point the same way: when energy dominates the cost of AI, the countries with cheap power and permission to build will host the industry. Now China has a golden opportunity to secure a strong lead in artificial intelligence.
If anyone still has doubts, AI is real and it works. I use AI extensively in my day-to-day work. The following is a real-time video of me using AI to code a simple game in half an hour. Without AI assistance it would have taken me four hours, so on this occasion AI made me eight times as productive.
Software development is not an isolated case. A recent MIT study found that just under 12% of US jobs can already be replaced by AI. I expect this to soar to 30% of all jobs by 2040.
MIT study finds AI can already replace 11.7% of U.S. workforce
PUBLISHED WED, NOV 26 2025 10:00 AM EST
UPDATED WED, NOV 26 2025 12:16 PM EST
MacKenzie Sigalos @KENZIESIGALOSKEY POINTS
- Massachusetts Institute of Technology released a study that found that artificial intelligence can already replace 11.7% of the U.S. labor market.
- The study was conducted using a labor simulation tool called the Iceberg Index, which was created by MIT and Oak Ridge National Laboratory.
- For lawmakers preparing billion-dollar reskilling and training investments, the index offers a detailed map of where disruption is forming down to the zip code.
Massachusetts Institute of Technology on Wednesday released a study that found that artificial intelligence can already replace 11.7% of the U.S. labor market, or as much as $1.2 trillion in wages across finance, health care and professional services.
The study was conducted using a labor simulation tool called the Iceberg Index, which was created by MIT and Oak Ridge National Laboratory. The index simulates how 151 million U.S. workers interact across the country and how they are affected by AI and corresponding policy.
…
Read more: https://www.cnbc.com/2025/11/26/mit-study-finds-ai-can-already-replace-11point7percent-of-us-workforce.html
See the MIT study here.
Not every industry is as affected by AI as software development currently is, but make no mistake: the AI industry has its pedal to the metal, trying to expand the range of work it can replace. Right now my industry feeds are full of advertisements for lawyers willing to train artificial intelligence legal advisors in return for obscene amounts of money. Law and accountancy may be the next professions to fall.
This isn’t the end of jobs. Plenty of jobs will resist an AI takeover for a long time, especially anything involving complex, unpredictable situations and a lack of accessible training data. But AI is an industrial revolution on a par with the dawn of the machine age in the late 1700s and early 1800s. Many jobs we take for granted today will no longer exist in a few years.
There will still be jobs, just maybe not the job you currently do.
Plenty of issues other than AI shaped the midterm result. And the impact of data centers on the people living near them has been handled badly. AI’s demand for electricity and water drives up local costs if those needs are not properly addressed when the data center is built.
Personally, I don’t want to work with Chinese AI. I don’t trust China to protect my privacy or respect my business confidentiality. China has a troubling national security law which requires private businesses to surrender data to the Communist Party on demand, without significant due process. That is an obvious opportunity for corruption, if seized data is then sold by corrupt officials to other parties.
But if a shift in US policy drives up the cost of US AI, many businesses won’t have a choice. If your competitors are operating on a lower cost base because they do business with China, the choices are to do as your competitors do, or start talking to your lawyers about Chapter 11.
My day job, software development, is likely to be one of the first professions to fall. Right now AI functions as a very capable junior programmer. It’s like having a team of low-cost programmers on call whenever I need software built. When AI moves up the capability ladder, which might be months or years from now, I’ll need to find something else to do. I can see what is coming, so I’m planning for it.
But restricting the construction of data centers in one country will not prevent the advance of AI. All it can do is ensure the benefits of hosting AI go to someone else, someone who, by hosting the AI your nation’s businesses are rapidly coming to depend on, will have the power to shut down much of your economy with the flick of a switch. Let’s hope the winner of that particular economic arms race is not China.
