Every time someone uses ChatGPT to write an essay, create an image or advise them on planning their day, the environment pays a price.

A query on the chatbot that uses artificial intelligence is estimated to require at least 10 times more electricity than a standard search on Google.

If all Google searches similarly used generative AI, they might consume as much electricity as a country the size of Ireland, calculates Alex de Vries, the founder of Digiconomist, a website that aims to expose the unintended consequences of digital trends.

Yet someone using ChatGPT or another artificial intelligence application has no way of knowing how much power their questions will consume as they are processed in the tech companies’ enormous data centers.

De Vries said the skyrocketing energy demand of AI technologies will no doubt require the world to burn more climate-warming oil, gas and coal.

“Even if we manage to feed AI with renewables, we have to realize those are limited in supply, so we’ll be using more fossil fuels elsewhere,” he said. “The ultimate outcome of this is more carbon emissions.”

AI is also thirsty for water. ChatGPT gulps roughly a 16-ounce bottle in as few as 10 queries, calculates Shaolei Ren, associate professor of electrical and computer engineering at UC Riverside, and his colleagues.

The increasing consumption of energy and water by AI has raised concerns in California and around the globe. Experts have detailed how it could stall the transition to green energy, while increasing consumer’s electric bills and the risk of blackouts.

To try to prevent those consequences, De Vries, Ren and other experts are calling on the tech companies to disclose to users how much power and water their queries will consume.

“I think the first step is to have more transparency,” Ren said. The AI developers, he said, “tend to be secretive about their energy usage and their water consumption.”

Ren said users should be told on the websites where they are asked to type in their queries how much energy and water their requests will require. He said this would be similar to how Google now tells people searching for airline flights how much carbon emissions the trip will generate.

“If we had that knowledge,” he said, “then we could make more informed decisions.”

Data centers — enormous warehouses of computer servers that support the internet — have long been big power users. But the specialized computer chips required for generative AI use far more electricity because they are designed to read through vast amounts of data.

The new chips also generate so much heat that even more power and water is needed to keep them cool.

Even though the benefits and risks of AI aren’t yet fully known, companies are increasingly incorporating the technology into existing products.

In May, for example, Google announced that it was adding what it called “AI Overviews” to its search engine. Whenever someone now types a question into Google search, the company’s AI generates an answer from the search results, which is highlighted at the top.

Not all of Google’s AI-generated answers have been correct, including when it told a user to add Elmer’s glue to pizza sauce to keep cheese from sliding off the crust.

But searchers who don’t want those AI-generated answers or want to avoid the extra use of power and water can’t turn off the feature.

“Right now, the user doesn’t have the option to opt out,” Ren said.

Google did not respond to questions from The Times.

OpenAI, the company that created ChatGPT, responded with a prepared statement, but declined to answer specific questions, such as how much power and water the chatbot used.

“AI can be energy-intensive and that’s why we are constantly working to improve efficiencies,” OpenAI said. “We carefully consider the best use of our computing power and support our partners’ efforts to achieve their sustainability goals. We also believe that AI can play a key role in accelerating scientific progress in the discovery of climate solutions.”

Three years ago, Google vowed to reach “net-zero” — where its emissions of greenhouse gases would be equal to what it removed — by 2030.

The company isn’t making progress toward that goal. In 2023, its total carbon emissions increased by 13%, the company disclosed in a July report. Since 2019, its emissions are up 48%.

“As we further integrate AI into our products, reducing emissions may be challenging due to increasing energy demands from the greater intensity of AI compute,” the company said in the report.

Google added that it expects its emissions to continue to rise before dropping sometime in the future. It did not say when that may be.

The company also disclosed that its data centers consumed 6.1 billion gallons of water in 2023 — 17% more than the year before.

“We’re committed to developing AI responsibly by working to address its environmental footprint,” the report said.

De Vries said he was disappointed Google had not disclosed in the report how much AI was adding to its power needs. The company said in the report that such a “distinction between AI and other workloads” would “not be meaningful.”

By not separately reporting the power use of AI, he said, it is impossible to calculate just how much more electricity Google search was now using with the addition of AI Overviews.

“While capable of delivering the required info,” he said, “they are now withholding it.”

The internet is becoming awash in words and images generated by artificial intelligence.

Sam Altman, OpenAI’s chief executive, wrote in February that the company generated about 100 billion words per day — a million novels’ worth of text, every day, an unknown share of which finds its way onto the internet.

A.I.-generated text may show up as a restaurant review, a dating profile or a social media post. And it may show up as a news article, too: NewsGuard, a group that tracks online misinformation, recently identified over a thousand websites that churn out error-prone A.I.-generated news articles.

In reality, with no foolproof methods to detect this kind of content, much will simply remain undetected.

All this A.I.-generated information can make it harder for us to know what’s real. And it also poses a problem for A.I. companies. As they trawl the web for new data to train their next models on — an increasingly challenging task — they’re likely to ingest some of their own A.I.-generated content, creating an unintentional feedback loop in which what was once the output from one A.I. becomes the input for another.

In the long run, this cycle may pose a threat to A.I. itself. Research has shown that when generative A.I. is trained on a lot of its own output, it can get a lot worse.

Here’s a simple illustration of what happens when an A.I. system is trained on its own output, over and over again:

While this is a simplified example, it illustrates a problem on the horizon.

Imagine a medical-advice chatbot that lists fewer diseases that match your symptoms, because it was trained on a narrower spectrum of medical knowledge generated by previous chatbots. Or an A.I. history tutor that ingests A.I.-generated propaganda and can no longer separate fact from fiction.

Just as a copy of a copy can drift away from the original, when generative A.I. is trained on its own content, its output can also drift away from reality, growing further apart from the original data that it was intended to imitate.

In a paper published last month in the journal Nature, a group of researchers in Britain and Canada showed how this process results in a narrower range of A.I. output over time — an early stage of what they called “model collapse.”

The eroding digits we just saw show this collapse. When untethered from human input, the A.I. output dropped in quality (the digits became blurry) and in diversity (they grew similar).

If only some of the training data were A.I.-generated, the decline would be slower or more subtle. But it would still occur, researchers say, unless the synthetic data was complemented with a lot of new, real data.

Degenerative A.I.

In one example, the researchers trained a large language model on its own sentences over and over again, asking it to complete the same prompt after each round.

When they asked the A.I. to complete a sentence that started with “To cook a turkey for Thanksgiving, you…,” at first, it responded like this:

“The model becomes poisoned with its own projection of reality,” the researchers wrote of this phenomenon.

This problem isn’t just confined to text. Another team of researchers at Rice University studied what would happen when the kinds of A.I. that generate images are repeatedly trained on their own output — a problem that could already be occurring as A.I.-generated images flood the web.

They found that glitches and image artifacts started to build up in the A.I.’s output, eventually producing distorted images with wrinkled patterns and mangled fingers.

“You’re kind of drifting into parts of the space that are like a no-fly zone,” said Richard Baraniuk, a professor who led the research on A.I. image models.

The researchers found that the only way to stave off this problem was to ensure that the A.I. was also trained on a sufficient supply of new, real data.

While selfies are certainly not in short supply on the internet, there could be categories of images where A.I. output outnumbers genuine data, they said.

For example, A.I.-generated images in the style of van Gogh could outnumber actual photographs of van Gogh paintings in A.I.’s training data, and this may lead to errors and distortions down the road. (Early signs of this problem will be hard to detect because the leading A.I. models are closed to outside scrutiny, the researchers said.)

Why collapse happens

All of these problems arise because A.I.-generated data is often a poor substitute for the real thing.

This is sometimes easy to see, like when chatbots state absurd facts or when A.I.-generated hands have too many fingers.

But the differences that lead to model collapse aren’t necessarily obvious — and they can be difficult to detect.

When generative A.I. is “trained” on vast amounts of data, what’s really happening under the hood is that it is assembling a statistical distribution — a set of probabilities that predicts the next word in a sentence, or the pixels in a picture.

For example, when we trained an A.I. to imitate handwritten digits, its output could be arranged into a statistical distribution that looks like this:

The peak of this bell-shaped curve represents the most probable A.I. output — in this case, the most typical A.I.-generated digits. The tail ends describe output that is less common.

Notice that when the model was trained on human data, it had a healthy spread of possible outputs, which you can see in the width of the curve above.

But after it was trained on its own output, this is what happened to the curve:

It gets taller and narrower. As a result, the model becomes more and more likely to produce a smaller range of output, and the output can drift away from the original data.

Meanwhile, the tail ends of the curve — which contain the rare, unusual or surprising outcomes — fade away.

This is a telltale sign of model collapse: Rare data becomes even rarer.

If this process went unchecked, the curve would eventually become a spike:

This was when all of the digits became identical, and the model completely collapsed.

Why it matters

This doesn’t mean generative A.I. will grind to a halt anytime soon.

The companies that make these tools are aware of these problems, and they will notice if their A.I. systems start to deteriorate in quality.

But it may slow things down. As existing sources of data dry up or become contaminated with A.I. “slop,” researchers say it makes it harder for newcomers to compete.

A.I.-generated words and images are already beginning to flood social media and the wider web. They’re even hiding in some of the data sets used to train A.I., the Rice researchers found.

“The web is becoming increasingly a dangerous place to look for your data,” said Sina Alemohammad, a graduate student at Rice who studied how A.I. contamination affects image models.

Big players will be affected, too. Computer scientists at N.Y.U. found that when there is a lot of A.I.-generated content in the training data, it takes more computing power to train A.I. — which translates into more energy and more money.

“Models won’t scale anymore as they should be scaling,” said ​​Julia Kempe, the N.Y.U. professor who led this work.

The leading A.I. models already cost tens to hundreds of millions of dollars to train, and they consume staggering amounts of energy, so this can be a sizable problem.

‘A hidden danger’

Finally, there’s another threat posed by even the early stages of collapse: an erosion of diversity.

And it’s an outcome that could become more likely as companies try to avoid the glitches and “hallucinations” that often occur with A.I. data.

This is easiest to see when the data matches a form of diversity that we can visually recognize — people’s faces:

This drop in diversity is “a hidden danger,” Mr. Alemohammad said. “You might just ignore it and then you don’t understand it until it’s too late.”

Just as with the digits, the changes are clearest when most of the data is A.I.-generated. With a more realistic mix of real and synthetic data, the decline would be more gradual.

But the problem is relevant to the real world, the researchers said, and will inevitably occur unless A.I. companies go out of their way to avoid their own output.

Related research shows that when A.I. language models are trained on their own words, their vocabulary shrinks and their sentences become less varied in their grammatical structure — a loss of “linguistic diversity.”

And studies have found that this process can amplify biases in the data and is more likely to erase data pertaining to minorities.

Ways out

Perhaps the biggest takeaway of this research is that high-quality, diverse data is valuable and hard for computers to emulate.

One solution, then, is for A.I. companies to pay for this data instead of scooping it up from the internet, ensuring both human origin and high quality.

OpenAI and Google have made deals with some publishers or websites to use their data to improve A.I. (The New York Times sued OpenAI and Microsoft last year, alleging copyright infringement. OpenAI and Microsoft say their use of the content is considered fair use under copyright law.)

Better ways to detect A.I. output would also help mitigate these problems.

Google and OpenAI are working on A.I. “watermarking” tools, which introduce hidden patterns that can be used to identify A.I.-generated images and text.

But watermarking text is challenging, researchers say, because these watermarks can’t always be reliably detected and can easily be subverted (they may not survive being translated into another language, for example).

A.I. slop is not the only reason that companies may need to be wary of synthetic data. Another problem is that there are only so many words on the internet.

Some experts estimate that the largest A.I. models have been trained on a few percent of the available pool of text on the internet. They project that these models may run out of public data to sustain their current pace of growth within a decade.

“These models are so enormous that the entire internet of images or conversations is somehow close to being not enough,” Professor Baraniuk said.

To meet their growing data needs, some companies are considering using today’s A.I. models to generate data to train tomorrow’s models. But researchers say this can lead to unintended consequences (such as the drop in quality or diversity that we saw above).

There are certain contexts where synthetic data can help A.I.s learn — for example, when output from a larger A.I. model is used to train a smaller one, or when the correct answer can be verified, like the solution to a math problem or the best strategies in games like chess or Go.

And new research suggests that when humans curate synthetic data (for example, by ranking A.I. answers and choosing the best one), it can alleviate some of the problems of collapse.

Companies are already spending a lot on curating data, Professor Kempe said, and she believes this will become even more important as they learn about the problems of synthetic data.

But for now, there’s no replacement for the real thing.