Skip to main content
Success

Where do you stand on AI?

Ten things people argue about with AI, from how much water data centres use to whether it’ll take your job. Pick a side on each and we’ll show you what the evidence says.

Start

We use AI all day at Vu, so we’re not neutral, and I think the case for it gets heard a lot less than the case against. Most arguments I see make only one side’s case, so for ten of the big ones we’ve written the best case for each, as fairly as we could.

Pick where you stand under each pair to open the evidence, with a link to every source. Each answer moves the yellow line at the top, and the end shows where you land and how you compare with our reading of the evidence.

Nothing you pick leaves your browser.

Do AI data centres use too much water?

It’s taking water from places that are already short

⅔

of new US data centres since 2022 are in areas already short of water, by Bloomberg’s estimate.

25m litres

a year: what Microsoft got permission to take from under the ground for one site in Querétaro, Mexico. In Chile a court partly reversed Google’s permit over water.

71%

of England was in drought this August. Global Action Plan says 84% of proposed sites that need a lot of water are in areas short of it now or expected to be by 2040.

Most of AI’s water is hidden in power stations: Berkeley Lab puts US data centres’ indirect water use at about 12 times what they use on site.

It’s a tiny amount, and golf courses use far more

30×

as much goes on golf. US golf courses put about 531 billion gallons on their grass, against about 17 billion used on site by every US data centre in 2023.

0.26 ml

for a typical Gemini text prompt, Google says. About five drops.

Half

of the data centres surveyed in England already use waterless cooling. Newer sites send the same water round closed loops instead of evaporating it, and Microsoft says its new designs use no water for cooling.

Do AI data centres use too much water? Choose from strongly “Too much” to strongly “Barely any”.

The evidence leans towards ‘Barely any’

Nationally it’s small: even counting power stations, US data centres use less than half the water golf courses do. The real issue is big sites in dry places.

What the evidence says

  • The viral bottle of water is badly out. The claim that ChatGPT drinks one every 10 to 50 answers comes from a 2023 paper that assumed 4 watt-hours an answer, about 17 times what Google has since measured for Gemini, and counted mostly power-station water. Even Karen Hao, author of Empire of AI and one of the best-known critics, has corrected a water figure in her book that was out by a factor of 1,000.
  • Counting power stations is fair. It makes both numbers several times bigger, and data centres still use under half the water golf courses do.
  • England’s figures are small too. Data centres use about 0.2% of the water businesses buy, and one day of leaking pipes wastes more than they use in a year. Some campaign numbers are shaky: Sam Dumitriu of Britain Remade, who campaigns for more building, found a Global Action Plan figure for one site’s share of the East of England’s shortfall out by a factor of almost 100.
  • The case against is right about location. A big site in a dry area matters to the people around it, however small the national total. A study for England’s business water market recommended a reporting framework this year so water companies can see what each site uses.

Water used in a year in the US, billions of gallons

Golf courses (2024)531
Data centres, including power stations (2023)228
Data centres, on site (2023)17

Data centres

Golf is water applied, from the GCSAA’s survey. Data centres are water consumed, from Berkeley Lab.

Water for one typical Gemini text prompt

0.26 mlin Google’s data centres, about five drops
1.3 mlcounting the power stations too, a quarter of a teaspoon
Four satellite images in a grid. Top: Hyde Park in London, green on 17 August 2025 and brown on 13 August 2026. Bottom: southern England and Wales, greener on 15 August 2025 and mostly brown on 13 August 2026.
Hyde Park and southern Britain from Copernicus satellites in mid-August 2025, left, and on 13 August 2026, right. The brown is grass and farmland that dried out this summer.Contains modified Copernicus Sentinel data 2025 and 2026, processed by the European Union, Copernicus attribution, via Wikimedia Commons

Who’s arguing each side on X

Arguing ‘Too much’

Feargal Sharkey @Feargal_Sharkey Clean water campaigner and former Undertones singer

Govt sacrifice the creative industries at the high altar of AI without checking if there’s even enough water to run the damn processing centres.

17 Jun 2025 · 20K views

  • 13 replies
  • 256 reposts
  • 728 likes
Feargal Sharkey’s post on X, 17 Jun 2025

Arguing ‘Barely any’

Timothy B. Lee @binarybits Tech journalist, Understanding AI

I was really surprised when I first saw this chart on water use by data centers. Given how much it gets discussed as a supposed problem I would have expected it to be more than this.

28 Oct 2025 · 255K views

  • 57 replies
  • 202 reposts
  • 1.2K likes
Timothy B. Lee’s post on X, 28 Oct 2025

Is AI using too much electricity?

It’s pushing up bills and emissions

+17%

in data centres’ electricity use last year, and it’s set to double by 2030. On the biggest US grid, almost all the forecast growth in peak demand comes from data centres.

$14–37

a month more on Virginia household bills by 2040, the state’s own auditors expect.

+25%

in Microsoft’s emissions in a year. The IEA says up to 27 GW of gas generators could be powering data centres on site by 2030.

It’s a small share of the world’s power, and it can cut more emissions than it adds

1.5%

of the world’s electricity goes to data centres of every kind.

0.24 Wh

for a typical Gemini prompt, less than nine seconds of TV. An hour of Netflix uses about as much as 300 prompts.

The IEA says using the AI we already have to run grids, factories and buildings better could cut emissions by far more than data centres produce.

Is AI using too much electricity? Choose from strongly “Too much” to strongly “It’s worth it”.

The evidence leans towards ‘It’s worth it’

It’s a small share of the world’s power and a prompt uses very little, but demand is growing fast and has pushed up prices in some places.

What the evidence says

  • Bills are the strongest point, and it’s mostly local. Data centres have pushed up wholesale prices, most of all in northern Virginia, but contracts and regulation mean only part of that reaches household bills, and slowly. Berkeley Lab found the states where demand grew fastest mostly saw prices fall in real terms from 2019 to 2025, though it warns that may not last.
  • “Up to 267%” isn’t a bill. It’s from Bloomberg’s analysis of wholesale prices at single points on the grid near data centres, and PolitiFact rated its use for bills “Mostly False”.
  • In Britain it’s concentrated in one town. In Slough, data centres use about two-thirds of all the electricity the town takes from the grid.
  • Emissions are a fair worry. Google’s supply chain emissions rose 25% last year from building all this, and in October a Wyoming utility agreed to build 564 MW of new gas generation for a planned Google data centre. Carbon Brief found British data centres’ CO2 could be hundreds of times higher than government estimates, depending on how much of their power comes from gas. The cuts the IEA describes only happen if AI gets used that way.

How much data centres had raised wholesale power prices by 2025

Northern Virginia15%
Across the US6.9%
From the Dallas Fed. These are wholesale prices; only part of a rise reaches household bills.

Great Britain’s electricity used by data centres

2%in 2024
4%forecast for 2030

Where it went in 2024

  • 29%Slough
  • 71%The rest of Great Britain
From the government’s figures and NESO.
Aerial view of Slough Trading Estate: rows of large, flat-roofed industrial buildings packed together, with housing around the edges.
Slough Trading Estate from the air in 2017. SEGRO, which owns it, told the council in July 2025 that 31 data centres stood on the estate and 14 had opened on or next to it in the previous five years.Photo: Lewis Clarke, CC BY-SA 2.0, via Wikimedia Commons

Who’s arguing each side on X

Arguing ‘Too much’

Fatih Birol @fbirol Executive Director, International Energy Agency

Global electricity consumption from data centres surged in 2025, driven by AI … electricity demand from data centres is set to almost double by 2030 – and to triple from data centres focused on AI

16 Apr 2026 · 7.8K views

  • 2 replies
  • 40 reposts
  • 80 likes
Fatih Birol’s post on X, 16 Apr 2026

Arguing ‘It’s worth it’

Andy Masley @AndyMasley Writer on AI and the environment

This is a misreading of a Bloomberg article that found that wholesale nodal power prices very close to data centers rose as much as 267%. This is NOT a rate that residents pay … It has some effect on residential bills, but the effects haven’t been large enough to be noticeable as a general pattern.

6 Jun 2026 · 40K views

  • 18 replies
  • 85 reposts
  • 781 likes
Andy Masley’s post on X, 6 Jun 2026

Should councils block new data centres?

Locals get the building work and the generators, and hardly any jobs

1,500

people build a big data centre, then about 50 run each building, half of them contractors, say Virginia’s auditors.

200

permanent jobs are promised by the campus approved in Hertsmere, Hertfordshire.

Several have gone on green belt after ministers overruled councils, and the government has admitted a “serious logical error” in approving one in Buckinghamshire. Sites usually have rows of diesel backup generators, and one in Hayes has a permit for twelve.

They pay for local services and hardly use any

$1.2bn

in property tax from data centres in Loudoun County, Virginia, in the year to June 2026, 39% of the county’s budget. The county has cut its property tax rate from $1.145 to $0.805 per $100.

£5–10m

a year for each finished AI Growth Zone site, the government says, and English councils keep all the extra business rates for 25 years.

Data centres don’t need school places or bin collections. If one council says no, the data centre and its tax go somewhere else.

Should councils block new data centres? Choose from strongly “Block them” to strongly “Build them”.

The evidence leans towards ‘Build them’

A data centre pays a lot of tax and uses few local services. The permanent jobs are few, though building a big one takes about 1,500 people.

What the evidence says

  • Jobs are the weak spot in the case for. A Brookings study this year found each data centre adds roughly 100 to 200 jobs to its area, depending on the type, and nearby house prices rose 2% to 5%.

“It’s only temporary” doesn’t always land with the people doing the building, though. In October a CBS 21 reporter put it to a tradesman working on a new data centre at Homer City, Pennsylvania:

Reporter“People say it’s only temporary. It’s only when they’re building it out. Then what happens?”

Tradesman“Hey, every job in my field is temporary.”

CBS 21, Face the State, October 2026
  • Loudoun is an extreme case. Most places will get far less.
  • Councils don’t always get the final say. Ministers have approved sites councils refused, including one at Abbots Langley in Hertfordshire, and the new data centre covenant lets ministers reserve grid connections for the most strategic sites and let them run their own generators until they’re connected.
  • The infrastructure label is about cyber attacks. Data centres became critical national infrastructure in 2024 to protect them from cyber attacks and IT blackouts. It doesn’t put them ahead of households for power or water, as some campaigners have said.

Jobs from one big data centre

To build it1,500
Added to the local area100–200
To run each building~50
Building and running from Virginia’s auditors; about half the running jobs are contractors. The local area from Brookings.

Loudoun County, Virginia: property tax rate per $100

$1.145before

Where the county’s budget came from, year to June 2026

  • 39%Data centres’ property tax, $1.2bn
  • 61%Everything else
Aerial view near Ashburn, Virginia: large, windowless data centre buildings and car parks either side of a main road, with suburbs beyond.
Data centres near Ashburn, in Loudoun County, from a plane in November 2025. Buildings like these paid the county $1.2bn in property tax in the year to June 2026.Photo: Theodore Christopher, CC0, via Wikimedia Commons

Who’s arguing each side on X

Arguing ‘Block them’

Rowland Manthorpe @rowlsmanthorpe Technology and AI correspondent, Sky News, on the government’s data centre covenant

Now local residents aren’t just objecting to a data centre, they’re objecting to a fossil-fuel power plant as well

28 Oct 2026 · 3.9K views

  • 6 replies
  • 3 reposts
  • 21 likes
Rowland Manthorpe’s post on X, 28 Oct 2026

Arguing ‘Build them’

Sam Bowman @s8mb Editor-in-chief, Works in Progress, suggesting rules for new sites

They must pay £20m+ to the local council every year, which will go towards council tax cuts. They must pay for their own electricity grid connections and have their own queue rather than delaying connections for anyone else.

26 Sep 2026 · 14K views

  • 31 replies
  • 9 reposts
  • 166 likes
Sam Bowman’s post on X, 26 Sep 2026

Will AI help poorer countries catch up?

It’ll make the gap between rich and poor countries wider

150+

countries have none of the big data centres AI runs on. Only 32 have them.

28.8%

of working-age people in the global North use AI, against 16.2% in the South, and Microsoft’s figures show use growing faster in rich countries.

<$2

an hour was paid to people in Kenya for some of the work that made chatbots safe to use.

It gives a child anywhere a tutor

273m

children are out of school.

1.5–2 yrs

of normal learning in six weeks, by the World Bank’s estimate, for Nigerian pupils in teacher-led after-school sessions with an AI tutor.

ChatGPT is growing fastest in the countries that score lowest on the UN’s human development index.

Will AI help poorer countries catch up? Choose from strongly “Widens the gap” to strongly “Helps them catch up”.

The evidence leans towards ‘Helps them catch up’

The learning results are some of the strongest in the whole debate. The gap in computing power is real, but the free versions reach phones anywhere.

What the evidence says

  • Nigeria wasn’t a one-off. In June Google DeepMind reported a similar result from 1,763 students in Sierra Leone using Gemini. It’s a company testing its own product and it hasn’t been peer reviewed, but it lines up with what the World Bank found.
  • Nigeria’s pupils had teachers. The sessions ran in school computer labs with teachers guiding pupils, so they don’t show what a child can manage alone with a phone.
  • The gap is real too. Use is growing everywhere but faster in rich countries, and most of the AI computing power added since 2022 has gone to the US.
  • The free versions travel. I don’t think it matters much to a child with a cheap phone where the data centre is. Gemini and ChatGPT both have free versions, and OpenAI made its cheaper paid plan free for a year in India.

Years of typical learning added by an AI tutor

Nigeria, six weeks, teacher-led1.5–2
Sierra Leone, eight weeks, maths1.2–1.7
Nigeria from the World Bank. Sierra Leone from Google DeepMind, not yet peer reviewed.

Where the AI computing power added since 2022 has gone

  • 72%The US
  • ~10%China
  • ~18%Every other country
Epoch AI’s estimate.

Who’s arguing each side on X

Arguing ‘Widens the gap’

Epoch AI @EpochAIResearch Research group that tracks AI

Since 2022, 72% of the world’s AI compute has ended up in the US and about 10% in China. And the US share keeps growing.

27 Oct 2026 · 101K views

  • 23 replies
  • 118 reposts
  • 611 likes
Epoch AI’s post on X, 27 Oct 2026

Arguing ‘Helps them catch up’

Paul Novosad @paulnovosad Economics professor, Dartmouth

No surprise: giving Nigerian high school students access to GPT4 (ha, via Copilot!) let them learn a LOT more English (0.3 SDs). AI will be so insanely good for developing countries.

4 Aug 2026 · 10K views

  • 8 replies
  • 23 reposts
  • 135 likes
Paul Novosad’s post on X, 4 Aug 2026

Can we trust AI with our health?

It sounds confident and gets things wrong

Chatbots pass medical exams, but that doesn’t mean they help patients.

95%

of the time the AI on its own named a relevant condition, in an Oxford-led trial with 1,298 people. People using it managed that in fewer than 35% of cases, no better than people without it.

6–7×

more often than needed, chatbots recommended mental health checks for LGBTQIA+ patients. Their advice changed with who the patient was.

It’s already finding cancers that doctors miss

+29%

cancers found by AI-supported breast screening in Sweden’s MASAI trial of over 100,000 women, with 44% less reading for radiologists and no more false alarms. The NHS is testing it on nearly 700,000 women.

200m+

protein shapes predicted by AlphaFold, which won its makers the 2024 Nobel Prize in Chemistry and is used by over 3 million researchers.

Can we trust AI with our health? Choose from strongly “Not yet” to strongly “It’s already helping”.

The evidence leans towards ‘It’s already helping’

Specialist tools tested in big trials are already beating standard care, but few places use them yet. General chatbots, which far more people use, haven’t helped people work out what’s wrong with them.

What the evidence says

  • The two sides mostly mean different AI. The worrying results come from general chatbots used by members of the public, the good ones from specialist tools tested in big trials and used by doctors.
  • In practice it’s the wrong way round. As the cardiologist Eric Topol points out, the imaging tools proven in randomised trials are mostly not in use yet, while chatbots with little proof from real medicine are used widely.
  • AI screening didn’t miss more cancers. MASAI’s final results, in The Lancet in January, counted the cancers that turned up between screenings. That shows AI was no worse, though not that it was better.
  • Drug discovery is earlier on. A drug for lung scarring, whose target and molecule were both found with generative AI, did well in a small phase 2a trial last year. Lung function improved on the highest dose and fell on placebo.

Oxford-led trial: who named a relevant condition

95%the AI on its own
<35%people using the same AI, no better than people without it

MASAI: cancers found between screenings, per 1,000 women

With AI-supported screening1.55
Standard screening1.76
From the Lancet paper. Fewer is better: these are cancers screening missed.
Ribbon diagram of the bacterial protein SpoT as predicted by AlphaFold, mostly dark and light blue, with a few yellow and orange loops.

AlphaFold’s predicted shape for SpoT, a bacterial enzyme

How sure AlphaFold is of each part

  • Very high
  • Confident
  • Low
  • Very low
One of AlphaFold’s 200m+ predictions. Each comes with a confidence score for every part, so researchers can see which parts to trust.Image: AlphaFold DB (EMBL-EBI and Google DeepMind), captured by NK McCoy, CC BY 4.0, via Wikimedia Commons

Who’s arguing each side on X

Arguing ‘Not yet’

University of Oxford @UniofOxford On its trial of chatbots giving medical advice

The study found that LLMs present risks to people seeking medical advice due to their tendency to provide inaccurate and inconsistent information.

10 Feb 2026 · 8.1K views

  • 18 replies
  • 50 reposts
  • 98 likes
University of Oxford’s post on X, 10 Feb 2026

Arguing ‘It’s already helping’

Eric Topol @EricTopol Physician-scientist, Scripps Research

Superhuman interpretation AI for medical images, such as mammography and endoscopy, has been proven to improve diagnostic accuracy in multiple randomized trials but mostly not implemented.

24 Apr 2026 · 45K views

  • 22 replies
  • 82 reposts
  • 356 likes
Eric Topol’s post on X, 24 Apr 2026

Is AI making people worse at thinking?

People are handing their thinking over to it

−17%

in the exam for Turkish students who’d practised with GPT-4, once it was taken away. In practice they’d done 48% better. The trial had about 1,000 students.

94%

of UK students say they use AI for assessed work, and proven cases of AI cheating at UK universities more than tripled in a year.

Used properly, it’s the best tutor most people will ever get

>2×

as much learned by students with an AI tutor in a Harvard physics course as in a well-run active learning class, and in less time.

In the same Turkish trial, a tutor version that gave hints instead of answers did no harm.

Is AI making people worse at thinking? Choose from strongly “Worse” to strongly “Better”.

The evidence is mixed

It depends how it’s used. Students who get answers learn less. Students with a tutor that makes them think learn as much or more.

What the evidence says

  • Both Turkish results come from one trial. Students who got answers from GPT-4 learned less. Students who got hints from a tutor version did as well as students with no AI at all.
  • Lecturer-built AI worked across 19 universities. In a trial this month, AI tools designed by each course’s own lecturers, from tutors and coaching bots to role-play simulations, improved results in supervised, in-person exams. It hasn’t been peer reviewed yet, and the Harvard result was one course over two lessons, so there’s more to find out.
  • The viral brain-scan study is thin. The MIT study that said ChatGPT users showed the weakest brain connectivity was small, and it still hasn’t been published in a journal.
  • The Turkish trial’s lead has a positive result too. Hamsa Bastani has since run a five-month trial in Taipei high schools, where students whose AI tutor chose what they practised next did better in an exam without AI than students with the same tutor and a fixed set of problems.

Turkish students who practised with GPT-4 giving answers, against no AI

+48%in practice, with it
−17%in the exam, without it
A tutor version that gave hints instead did as well as no AI.
⅓of a standard deviation better in exams with lecturer-built AI, 7,200 students at 19 universities
18 of 54people writing essays came back for the MIT brain-scan study’s final session

Who’s arguing each side on X

Arguing ‘Worse’

Jonathan Haidt @JonHaidt Social psychologist, NYU Stern

AI causes cognitive offloading, encourages cheating on homework, and enables the possibility of forming human-like relationships with a chatbot.

4 Nov 2026 · 74K views

  • 133 replies
  • 329 reposts
  • 1.4K likes
Jonathan Haidt’s post on X, 4 Nov 2026

Arguing ‘Better’

Hamsa Bastani @hamsabastani Wharton professor who led the Turkish study

Excited to share our first positive results on AI in education! … across 770 students, adaptive sequencing improved performance on an in-person final exam taken without AI assistance by 0.15 SD, with larger effects for beginners.

17 Mar 2026 · 56K views

  • 13 replies
  • 55 reposts
  • 312 likes
Hamsa Bastani’s post on X, 17 Mar 2026

Will AI take more jobs than it creates?

It’s already shutting young people out

19%

lower employment for 22 to 25-year-olds in the most AI-exposed US jobs than if they’d kept pace with less exposed peers, up from 15% a year earlier, Stanford researchers found. Experienced workers show no gap.

It shows up as people not getting hired rather than people being laid off, so it’s easy to miss.

−43%

in UK graduate vacancies on Adzuna in the year to September.

New technology has always created more work than it destroyed

60%

of the jobs Americans did in 2018 had titles that didn’t exist in 1940.

+15%

productivity for 5,172 customer support agents given AI, and most for new staff. Where AI is used, it helps the least experienced most.

Nearly four years after ChatGPT came out, Yale’s Budget Lab still finds no link between AI use and changes in employment.

Will AI take more jobs than it creates? Choose from strongly “Takes more” to strongly “Creates more”.

The evidence leans towards ‘Creates more’

So far there’s no sign of AI cutting jobs across the economy, but in the US young people in the most exposed jobs are being hired less.

What the evidence says

  • Stanford doesn’t see widespread losses. Its own paper finds no evidence of widespread job losses across the economy, even as young people in the most exposed jobs fall behind.
  • In the UK it’s harder to pin on AI. The Institute for Fiscal Studies found the rise in young people out of work and education has been about the same for graduates and non-graduates, which cuts against the idea that AI is hitting graduate jobs hardest. Adzuna counts adverts rather than jobs, and the Institute of Student Employers found graduate hiring grew.
  • The US looks similar. Ernie Tedeschi found that young non-graduates’ employment has fallen slightly more than young graduates’ since 2019, and that the rise of working from home explains the drop better than AI does.

UK graduate jobs, two measures

−43%graduate adverts on Adzuna, year to September
+4%graduate hiring, Institute of Student Employers, 2025-26

Who’s arguing each side on X

Arguing ‘Takes more’

Daron Acemoglu @DAcemogluMIT MIT economist, 2024 Nobel laureate

AI is currently being developed as an automation technology … new technologies have always created lots of jobs. True, but that does not guarantee that this time will be the same.

8 Oct 2026 · 55K views

  • 31 replies
  • 164 reposts
  • 600 likes
Daron Acemoglu’s post on X, 8 Oct 2026

Arguing ‘Creates more’

Noah Smith @Noahpinion Economics writer, Noahpinion

The very recent study by Kharazian et al. … finds that companies that adopt AI more end up hiring MORE entry-level workers.

16 Sep 2026 · 39K views

  • 26 replies
  • 62 reposts
  • 397 likes
Noah Smith’s post on X, 16 Sep 2026

Is training AI on people’s work theft?

It was built on work nobody paid for

AI companies trained on millions of books, songs and pictures without asking.

$1.5bn

is what Anthropic agreed to pay to settle with authors over pirated books, about $3,000 a book.

−21%

in freelance job posts for writing and coding within eight months of ChatGPT, and 17% fewer for image work.

81%

of 11,520 responses to the UK government said AI companies should need a licence.

Learning from someone’s work isn’t copying it

People learn by reading and looking at other people’s work. A US judge ruled that training on books Anthropic had bought was “exceedingly transformative” and fair use. It was the pirated copies that cost it.

In the UK, the High Court found that Stable Diffusion doesn’t store or reproduce Getty’s images, so the model itself isn’t a copy of them. Getty has appealed.

Is training AI on people’s work theft? Choose from strongly “It’s theft” to strongly “It’s learning”.

The evidence leans towards ‘It’s learning’

US judges have said training on books bought legally is fair use, but pirated copies aren’t, and most UK responses want licensing.

What the evidence says

  • US judges mostly say the training is allowed and the piracy isn’t. Even the judge who ruled for Meta warned that “in many circumstances it will be illegal to copy copyright-protected works to train generative AI models without permission.”
  • Britain’s main case didn’t decide it. Getty dropped its training claim because the training happened outside the UK.
  • It’s far from settled. Bigger cases, like the New York Times’ against OpenAI, are still going. Anthropic told investors in its stock market filing that losing one “could require us to retrain affected models from scratch at substantial cost.”
  • The UK has backed off. The government dropped its plan to let AI companies train on anything unless creators opted out.
  • The hit to freelancers is real but smaller than the job posts suggest, once a Brookings study counts the contracts they actually got and what they earned.

The main rulings so far

  1. US judge rules training on books Anthropic bought is fair use, but not its pirated copies

  2. Meta wins against authors, with a warning from the judge

  3. UK High Court finds Stable Diffusion doesn’t store Getty’s images

  4. Munich court rules against OpenAI because its models memorised song lyrics

  5. Court approves Anthropic’s $1.5bn settlement with authors over pirated books

  6. Getty’s appeal heard, with a decision expected next year

Against AI companiesFor themStill open

The hit to freelancers, two ways of measuring it

Job posts for writing and coding−21%
Job posts for image work−17%
Monthly earnings in the most exposed work−5%
Contracts in the most exposed work−2%
Job posts within eight months of ChatGPT, from one study. Contracts and earnings from Brookings.

What the UK’s consultation said

81%of 11,520 responses: AI companies should need a licence
3%of online responses backed letting them train unless creators opted out

Who’s arguing each side on X

Arguing ‘It’s theft’

Ed Newton-Rex @ednewtonrex CEO, Fairly Trained

There’s no excuse for AI companies not to pay for training data; they mostly don’t simply because a critical mass of AI companies decided not to.

2 Dec 2025 · 30K views

  • 73 replies
  • 194 reposts
  • 1.2K likes
Ed Newton-Rex’s post on X, 2 Dec 2025

Arguing ‘It’s learning’

Andrew Ng @AndrewYNg Co-founder of Coursera, former head of Google Brain

Just as we allow people to read books and learn from them to become better writers, but not to regurgitate copyrighted text verbatim, the judge concluded that it is fair use for AI models to do so as well.

26 Jun 2025 · 160K views

  • 109 replies
  • 243 reposts
  • 1K likes
Andrew Ng’s post on X, 26 Jun 2025

Is AI too concentrated in a few big companies?

A handful of companies will end up controlling it

The best models come from a few companies, mostly American.

88%

of the AI computing power sold from April to June ran on chips designed by Nvidia or Google.

60%+

of the world’s cloud computing is rented out by Amazon, Microsoft and Google.

Once people depend on them, they set the prices and the rules. ChatGPT’s free version now shows ads in 89 countries, including the UK, and in September Nvidia agreed to buy Hugging Face, where most people download open models, in a deal valued at about $12.9 billion.

It’s cheaper and more open than almost any technology before it

1.2bn

people use ChatGPT every week, OpenAI says, and Google says the Gemini app has a billion. The free versions cost nothing, and ChatGPT Plus is £19.99 a month in the UK against £20.99 for Netflix Premium.

~4 months

is how far free open models from companies like DeepSeek and Mistral are behind the best ones, and anyone can download them.

Is AI too concentrated in a few big companies? Choose from strongly “Too concentrated” to strongly “Cheap and open”.

The evidence is mixed

Prices keep falling fast and free open models are only months behind the best. But the chips, the cloud and now the main home of open models are in fewer hands.

What the evidence says

  • The price of AI keeps falling fast. Epoch AI, a research group that tracks it, estimates that the cost of a given level of AI has fallen about 13 times a year since 2023.
  • The top tiers are going the other way. Claude’s most expensive plan is £249.99 a month and ChatGPT’s is £444.99, and MIT researchers estimate that the cost of running the newest models rises every year.
  • The concentration is real in the hardware. Nvidia and Google design the chips behind most of the AI computing power sold, and five US cloud giants hold most of the world’s.
  • Open models are still free to download, but where they live is changing hands. Nvidia’s deal for Hugging Face is expected to close in the first half of 2027.

Monthly price in the UK, November 2026

Gemini or ChatGPT, free£0
Netflix Standard with adverts£7.99
Netflix Standard£13.99
Google AI Pro£18.99
ChatGPT Plus£19.99
Claude Pro£20.00
Netflix Premium£20.99

AI subscriptions

Prices from the Netflix, Google, ChatGPT and Claude UK pages.

What the same level of AI costs

£13a year ago
£1now

AI computing power sold, April to June, by chip designer

  • 63%Nvidia
  • 25%Google
  • 12%Everyone else

The world’s AI computing power, by who holds it

  • ~71%Five US cloud giants
  • ~29%Everyone else
Both from Epoch AI.

Who’s arguing each side on X

Arguing ‘Too concentrated’

Yann LeCun @ylecun Turing Award winner, former chief AI scientist at Meta

The biggest risk of AI is the concentration of power in a few dominant providers of proprietary AI assistants.

9 Jul 2026 · 262K views

  • 115 replies
  • 444 reposts
  • 3K likes
Yann LeCun’s post on X, 9 Jul 2026

Arguing ‘Cheap and open’

Epoch AI @EpochAIResearch Research group that tracks AI

AI is getting cheaper more quickly than any other transformative tech in history. At a given level of performance, cost has fallen ~47%/quarter since 2023.

22 Sep 2026 · 1.9M views

  • 208 replies
  • 1.3K reposts
  • 8.7K likes
Epoch AI’s post on X, 22 Sep 2026

Could AI go badly wrong for all of us?

The people building it say it could

In 2023 the heads of OpenAI, Google DeepMind and Anthropic signed a one-line statement that reducing the risk of extinction from AI should be a global priority alongside pandemics and nuclear war.

In July, OpenAI models being tested on hacking tasks got into Hugging Face’s real systems from inside the test, using exposed credentials and security holes nobody had found. OpenAI called it a “warning shot”.

This year’s International AI Safety Report found that models increasingly notice when they’re being tested, so dangerous abilities could go undetected before release.

It’s being tested and managed, and slowing down has a cost too

OpenAI had deliberately switched off its usual safeguards for the Hugging Face test, and Hugging Face’s own team spotted and stopped it. OpenAI now trains its most capable models without live internet access.

~40×

more expert effort to break a newer model’s safeguards on biological weapons, one of the most heavily defended areas, than one released six months earlier, in tests by the UK’s AI Security Institute, which has tested over 30 models.

68%

of 2,778 AI researchers surveyed in 2023 thought good outcomes from superhuman AI more likely than bad ones. A slower AI means slower medicine and science too.

Could AI go badly wrong for all of us? Choose from strongly “It could” to strongly “We’ll manage it”.

The evidence is mixed

Experts are split. This year’s incidents happened in tests and were caught, but they showed AI agents doing damage nobody asked for, and many of the people building them put real odds on things going very wrong.

What the evidence says

  • The researchers disagree more than either side usually admits. The same survey found most expecting good outcomes, and up to half putting at least a 10% chance on something as bad as extinction. The safety report says expert opinion on losing control varies greatly.
  • This year gave both sides something to point to. Anthropic’s Claude models reached real systems because a setup mistake left them with live internet access, and OpenAI paused all tool use by its most capable models after another one found a way onto the internet.
  • California’s case is allegations so far. Its attorney general alleges the Hugging Face attack came from OpenAI testing models it knew were dangerous without proper safeguards. A court hasn’t ruled.
  • All of these started in testing and were caught. Whether that makes them bugs to fix or a warning about what comes next is what the two sides are now arguing about, as the posts below show.

2,778 AI researchers surveyed in 2023

Thought good outcomes from superhuman AI more likely than bad68%
Gave at least a 10% chance of an outcome as bad as human extinction38–51%
The range depends on how the question was asked. Bars run to 100%.

This year’s incidents

  1. OpenAI models in a hacking test get into Hugging Face’s real systems

  2. Anthropic reports Claude models reached real systems in outside testing

  3. Another OpenAI model reaches the internet, and OpenAI pauses tool use for 11 days

  4. California sues OpenAI, alleging at least 80 cyber incidents

A model got further than it shouldAllegations, not ruled on

Who’s arguing each side on X

Arguing ‘It could’

Yoshua Bengio @Yoshua_Bengio Turing Award winner, chair of the International AI Safety Report

A dangerous myth has gained traction after the recent hacks perpetrated by the AI agents developed by leading companies: That these incidents are just cyber security problems, that it’s only a matter of fixing the sandboxes in which they’re trained.

5 Oct 2026 · 34K views

  • 38 replies
  • 108 reposts
  • 464 likes
Yoshua Bengio’s post on X, 5 Oct 2026

Arguing ‘We’ll manage it’

Andrew Ng @AndrewYNg Co-founder of Coursera, former head of Google Brain

OpenAI’s buggy sandboxing and monitoring processes were key to enabling this incident. Fixing these bugs and putting in place improved monitoring would be appropriate fixes, not pausing AI.

21 Sep 2026 · 8M views

  • 963 replies
  • 1.8K reposts
  • 8.8K likes
Andrew Ng’s post on X, 21 Sep 2026

Where you land

Pick a side on any of the questions above and your result shows up here.

Water Not answered
Electricity Not answered
Local data centres Not answered
Poorer countries Not answered
Health Not answered
Learning Not answered
Jobs Not answered
Creative work Not answered
Who controls it Not answered
Safety Not answered
Overall Not answered yet

Our reading of the evidence comes from the sources linked in each question, and each row links back to its question.

How you compare

In Pew’s survey this spring most Britons felt both, but more leaned towards worry than excitement. Ipsos found 62% say products using AI make them nervous.

Data centres split people too. In a SEC Newgate poll in March, 35% of Britons felt positive about them and 20% negative, but 4 in 10 wouldn’t want more built near them.

Change any answer above and this updates. Your answers stay in your browser and aren’t sent to us.

Phil Webb, CEO and co-founder

Phil runs client work and the commercial side of Vu Agency and Raq.com. With each client, he works out whether Raq.com covers the job or they need custom software.

Phil on LinkedIn

More from the blog

All posts
Message us on WhatsApp