Home
로그인
회원가입
Americans Have Turned Against AI - Video học tiếng Anh
듣기 연습
듣기 연습
/
Video
/
The Infographics Show
/
Americans Have Turned Against AI
Americans Have Turned Against AI
학습 모드 선택:
자막 보기
단어 선택
단어 다시 쓰기
Highlight:
3000 Oxford Words
4000 IELTS Words
5000 Oxford Words
3000 Common Words
1000 TOEIC Words
5000 TOEFL Words
자막 (210)
0:00
Millions of Americans will open ChatGPT in the next 10 minutes…
0:03
And most of them believe it's making the world worse. Half the country uses
0:07
AI now. Yet only 16% of Americans think it will be good for society.
0:12
It’s the lowest level of optimism any major technology has launched with.
0:16
The people building AI call it humanity's greatest
0:18
breakthrough. The people using it describe something closer to a hostage situation.
0:23
So how did the fastest-adopted technology of our lifetime also become one of the most hated?
0:28
You might expect that it’s the older generations that are pushing back on
0:31
the AI revolution. After all, every new technology has faced resistance from
0:36
people who didn't grow up with it. But that's not what's happening.
0:39
The biggest users of AI are Americans under 30. They use it for schoolwork,
0:44
job applications, coding, brainstorming, and as a substitute for Google. By every
0:49
indicator, they should be AI's biggest fans. Instead, they're some of its biggest critics.
0:54
A Gallup study found the share of Gen Z who say AI makes them feel angry climbed from
0:58
22% to 31% by early 2026. And according to Pew Research, 40% of young Americans said it would
1:05
make the country worse than better. But this isn't just a Gen Z story.
1:09
Trust in AI is collapsing all across the U.S. Two thirds of
1:13
people think it's being developed too quickly. 59% don't trust the companies
1:17
building it. And most don't believe the government can keep it under control.
1:20
Yet people keep using it. ChatGPT alone is used by 44% of American adults, more than
1:27
double its share in 2023. That's the AI paradox.
1:31
So why does adoption keep accelerating if confidence keeps falling?
1:34
For most people, it wasn’t really adopted. Workplaces began implementing it in their
1:39
workflow. Schools added it to assignments. Apps and search engines displayed it as a main feature.
1:44
People had no other choice. It was forced upon them. AI didn’t earn their trust.
1:49
They have no other option.
1:51
The shift happened so gradually that most people barely noticed it.
1:54
Until suddenly, it was everywhere.
1:56
The moment you open Google and search for something, you’re hit with a block
2:00
of AI-written text answering the question before you even reach the search results.
2:04
The company that’s supposed to help you find information is now deciding what you see first.
2:09
And it works.
2:10
60% of Americans say they read the summaries regularly, because there’s no way to avoid
2:15
them. The moment Google started answering the question itself, the experience changed.
2:20
And the more the internet degraded, the less people trusted it. In 2021, 37% of Americans said
2:26
the spread of AI worried them more than it excited them. By 2025, that number had climbed to 50%,
2:32
even though most people hadn't been personally affected by AI in a professional capacity.
2:37
That would soon change.
2:39
In April 2025, Tobi Lütke, Shopify’s chief executive,
2:43
posted a memo to his staff. Before any team could ask him for a bigger budget or another employee,
2:47
they had to prove that a machine couldn’t do the work first. Lütke
2:51
insisted that “reflexive AI use was the new baseline expectation”.
2:55
Shopify wasn’t the only company going all in on AI.
2:59
Language learning platform Duolingo was next. CEO, Luis von Ahn, told his staff the company was going
3:04
“AI-first.” It would rely on AI before hiring, reduce the number of contractors, and weigh AI
3:10
use in performance reviews. Within a week, 148 new courses were shipped, it was work that would take
3:17
staff years to develop. The same AI implementation turned up at Box, at Fiverr, and other firms.
3:23
The wording barely changed. Run the software before you ask for new hires or extra money.
3:28
Then the pushback happened.
3:29
Longtime Duolingo users filmed themselves deleting the app in protest. Staff started asking if they
3:34
had to use AI well or just use it. Almost a year later Luis von Ahn walked his decision back.
3:40
He acknowledged that the AI-written code was hard to debug and the lessons weren’t reliable.
3:45
But the damage had been done.
3:47
And it was just the beginning.
3:48
The fear of AI taking jobs is real, so in the summer of 2025, a Stanford economist named Erik
3:54
Brynjolfsson decided to look into it. His team at the Digital Economy Lab pulled the records from
3:59
ADP, the firm that processes the payroll for millions of American workers, and studied who
4:04
still had a job, month after month. The AI impact was real.
4:09
Just now where people expected it.
4:11
The results showed that workers aged 22 to 25 in software and customer support had dropped
4:16
about 13% since late 2022. Workers over 30 in the same roles remained stable or
4:22
actually grew. It was the people just walking in the door that were at risk of an AI cull.
4:27
Some don’t get in the door at all.
4:29
Revelio Labs, which tracks hiring across the economy, counts roughly 35% fewer entry-level
4:35
postings since the start of 2023. Once employers saw what the best models could do,
4:40
many stopped graduate hiring. That freeze only surfaced in the payroll data a year or 2 later.
4:46
Brynjolfsson’s team found that when a company used AI to take over a task completely,
4:51
employment among younger workers fell. When AI was used to make a person faster,
4:55
employment remained stable, and sometimes grew.
4:58
Same technology, opposite result.
5:00
Everything comes down to one decision:
5:02
does a company use AI to replace a worker, or to back one up?
5:05
Near the bottom of the pay scale, replacement is winning.
5:08
Salesforce froze much of its junior hiring for a stretch and credited its
5:12
own AI for work graduates used to do. Goldman Sachs economists
5:16
found the results. The big names weren’t even pretending anymore.
5:19
But not every bet paid off.
5:21
Klarna, the buy-now-pay-later company, spent a year bragging that its chatbot did the work of
5:26
hundreds of agents and that it had stopped hiring people. They later admitted the bot gave “lower
5:32
quality” service and started hiring humans back. But some companies didn’t see that as a reason
5:36
to slow down. They saw it as a reason to aim lower… and the contractor was in the cross hairs.
5:42
An MIT report on AI in business found that AI is wiping out outsourced and contract roles,
5:47
not full-time staff. In mid-2025, the data-labeling firm Scale AI laid off 200
5:53
staff and cut ties with 500 contractors. That same report put the share of jobs
5:57
AI can already do at roughly 3% today, and saw it climbing over the longer run.
6:02
Graduates weren’t losing out to people with a better resume.
6:05
They were losing out period. There was not job.
6:09
By early 2026, nearly half of employed Gen Z said AI's risks at work now outweigh its benefits,
6:14
a jump of 11 percentage points in just one year. And when they're handed the same piece of work,
6:20
69% overwhelmingly trust the human version over the AI one. An entire generation now spends its
6:26
days training the technology that's replacing the jobs they're trying to build careers in.
6:31
The lost jobs are easy enough to measure. But the resentment isn't just about money.
6:35
In July 2024, all it took was a power surge across northern Virginia for nearly 60 data
6:40
centers to drop off the grid. Roughly 1,500 megawatts of demand vanished instantly.
6:45
Operators scrambled to get the surge under control before it took down anything else.
6:49
For a few seconds, the densest data-center hub in the country just wasn’t there.
6:54
Virginia holds more of these buildings than anywhere else,
6:57
and by 2023 they were already drawing about 26% of the state’s electricity. That’s why
7:02
one voltage swing could shock the whole grid. Each of those buildings runs hot. One question
7:08
to a chatbot burns about 2.9 watt-hours, that’s roughly 10 times a normal web search.
7:13
On its own, it’s nothing. But nobody asks just one question.
7:17
Multiply it by billions of questions a day, every day, in warehouses that never switch off, and it
7:22
becomes 183 terawatt-hours a year. That’s about 4% of all the power the country burns. That's
7:28
already 540 kilowatt-hours running in your name every year, whether you've ever touched a chatbot
7:34
or not. And the forecasts are only going up, as high as 17% of the nation's electricity by 2030.
7:41
More than 40% of the power fuelling the data centres comes from natural gas. Before the decade
7:46
is out, these buildings are set to burn more electricity than the entire country uses to make
7:50
its steel, cement, aluminum, and chemicals… Combined.
7:54
A high-voltage line takes up to 8 years to build, meaning a fifth of these data center projects are
7:59
about to run straight into a power shortage. To cover the expense of these extra lines,
8:04
utilities are going to regulators to demand more money from regular consumers. This means
8:09
your monthly expenses are going up by 2030 just to keep the lights on at a server farm down the road.
8:15
You didn't leave your lights on or crank up your AC, but your money
8:18
is bleeding out anyway to power the massive cooling fans of a tech giant.
8:22
Researchers at Carnegie Mellon estimate it could push the average American
8:26
power bill up about 8% by 2030, and in the worst-hit parts of Virginia, north of 25%.
8:33
There simply isn’t enough energy or money to fund it.
8:36
So, some companies got tired of waiting.
8:38
Instead of waiting years for the power grid to catch up, they built their own
8:42
power plants beside the data centers. And one company took that idea to the extreme.
8:47
To power Colossus, the supercomputer training Elon Musk's Grok chatbot,
8:51
xAI deployed a fleet of gas turbines in South Memphis.The entire site was up and
8:56
running in under 122 days, a project that would normally take years. By the time the
9:01
Southern Environmental Law Center flew thermal drones overhead, it counted around 35 turbines,
9:06
most of them without air permits. Just downwind sits Boxtown,
9:10
a neighborhood that has spent decades surrounded by refineries and chemical plants.
9:14
Residents began showing up at public hearings describing a rotten-egg smell in the air and
9:19
children wheezing through the night. Memphis already earns an "F" for ozone pollution from
9:24
the American Lung Association, and Boxtown's cancer risk is about 4 times the national average.
9:29
Then researchers at the University of Tennessee analyzed satellite data for TIME and found peak
9:34
nitrogen dioxide levels near the site had jumped roughly 79% compared with before xAI arrived.
9:41
The turbines were labelled as temporary,
9:43
a workaround that allowed them to keep running and bypass the regulations. Lawsuits followed,
9:48
but by the time the case moved forward, the servers had already been running for a year.
9:52
Power isn't the only resource these data centers consume.
9:55
In Newton County, Georgia, about an hour and a half east of Atlanta,
9:59
Beverly and Jeff Morris live 1,000 feet (305 meters) from a Meta data center. Their house
10:04
runs on a private well, which means their drinking water comes straight out of the
10:07
ground. It worked fine for years, right up until Meta broke ground next door in 2018.
10:13
A few months after construction began, the water pressure slowed to a trickle,
10:17
then stopped altogether. When it eventually came back, it was brown and murky. Beverly
10:22
Morris told the New York Times she was afraid to drink the water coming from her own tap.
10:26
The neighboring data center uses roughly 500,000 gallons every day, nearly a tenth
10:31
of all the water Newton County pumps. County projections show demand outpacing supply by 2030,
10:37
while residents face water bills expected to jump 33% in just 2 years,
10:41
compared with the usual 2% increase. A Meta funded study said the company’s operations
10:46
probably aren't to blame and has invested more than $4.5 million in local schools and nonprofits.
10:52
Private wells like the Morrises' fall outside the federal Safe Drinking Water Act,
10:57
so Meta didn’t break any laws. The Environmental Protection Agency is powerless, so Meta keeps
11:02
pulling water to cool the chips. And these hubs are insatiable.
11:07
A 2023 study by UC Riverside and the University of Texas found that training one model, GPT-3, ran
11:13
through about 5.4 million liters of fresh water… before the public ever typed a word. The same
11:20
researchers found that a short back-and-forth, around 100 words, consumes close to a full bottle.
11:25
One Google data center in Papillion, Nebraska, went through about 1.5 billion liters in 2024,
11:31
more than 600 Olympic pools, and it’s one site among thousands. That daily draw dwarfs the
11:37
one-time cost of training. Nationally, data centers used over 60 billion liters in 2023,
11:42
and the World Resources Institute expects AI to pass a trillion gallons a year by 2030.
11:47
In August 2025, Tucson’s city council voted 7 to 0 to reject Project Blue,
11:53
an Amazon-linked data-center campus that would have drained millions of
11:56
gallons of city water in the middle of a desert. Residents had packed the chamber
12:01
for weeks under a banner that read No Desert Data Center. For one night,
12:05
it looked like a regular city could tell a tech giant no, and actually make it stick.
12:10
It didn’t end there.
12:11
The developer already owned land just outside the city limits and kept building anyway. Then,
12:16
in the spring of 2026, the contractor was caught drawing enough Tucson water to supply
12:21
5 households for a year. The city shut it off. Mayor, Regina Romero, has been writing
12:26
ordinances to guard against the next data center, because there’s always a next one.
12:31
Tucson wasn’t alone.
12:32
In The Dalles, Oregon, a Google plant ended up using close to a quarter of the entire town’s
12:37
water. The company fought to keep those numbers hidden from the public through court orders.
12:41
Newer facilities can cool with air or closed-loop systems that use almost no
12:46
local water. Microsoft has already converted 2 Arizona sites to do exactly that. But the
12:51
biggest water users aren't the new builds, they're the older data centers. The ones
12:55
still sucking up millions of gallons instead of returning it to the river.
12:58
So while families shorten their showers and wait until dark to water the lawn,
13:02
the data center down the road keeps pulling water at full capacity, every hour, every night.
13:07
Jobs. Electricity. Water. Three separate battles happening in different parts of
13:11
the country. But they all lead back to the exact same business model.
13:14
In 2024 alone, Amazon, Microsoft, Google, and Meta spent more than $200 billion building the data
13:20
centers behind AI, up over 60% in a single year. Management consultancy firm McKinsey expects $5.2
13:27
trillion to be funnelled into them by 2030. That money buys the land, the power, and the water.
13:33
But there’s one more thing it buys. Something nobody signs off on: the writing, art,
13:38
and code scraped off the open internet to train the models in the first place.
13:42
To the companies, that’s an investment in the platform of the century. To the
13:46
creatives whose work fed it, it looks different. Their work was
13:50
harvested for free, then rented back to everyone by the month.
13:54
Not one of them was asked. Not one of them got paid.
13:57
Publishers and authors have sued in waves arguing their work was stolen and used to train AI without
14:02
permission. The courts still haven't decided where the legal line is, and so far none of
14:06
it has been ruled illegal. But AI is built on other people's writing, powered by other people's
14:12
electricity, cooled with other people's water… Then sold back to those same people as progress.
14:17
Now, Americans are starting to notice. The companies insist the costs are temporary.
14:21
Google and Microsoft both promise they'll return more water than they use by 2030 and keep shifting
14:27
to cleaner energy. But if you've already lost your job, your power bill just went up, or your
14:32
town's water is running low, a promise four years from now doesn't feel like much of an answer.
14:36
Between May 2024 and March 2025, that anger blocked or delayed about $64 billion in
14:43
data-center projects nationwide, according to Data Center Watch. In Virginia alone,
14:48
42 separate groups were working to slow or stop new ones.
14:51
What nobody agrees on is what to do about it.
14:54
No one is demanding an outright ban. No one is cheering it on.
14:58
There’s just a country caught between needing the tool and resenting it.
15:02
Most of these fights end the same way. A council denies the water,
15:05
the way Tucson did. The county next door sells the land. A state commission signs off on the
15:11
power. The building goes up anyway, trading its thirst for water for a heavier pull on the grid.
15:16
Residents keep learning the same lesson: the project can almost always route around them.
15:20
This was never about whether AI would arrive.That was inevitable. It was about
15:24
whether anyone still had the power to stop it. By the time we started asking that question,
15:29
the answer was already no. It’s not just ordinary Americans
15:32
who are losing faith in AI. The people building it are too. Watch Why AI Researchers Are Quitting
15:38
and Panicking on the Way Out to find out why the industry is collapsing. Or click on this video.
