Your AI prompts could have a hidden environmental cost
Each word in an AI prompt is broken down into clusters of numbers called 'token IDs' and sent to massive data centers — some larger than football fields — powered by coal or natural gas plants. There, stacks of large computers generate responses through dozens of rapid calculations.
The whole process can take up to 10 times more energy to complete than a regular Google search, according to a frequently cited estimation by the Electric Power Research Institute.
So, for each prompt you give AI, what's the damage? To find out, researchers in Germany tested 14 large language model (LLM) AI systems by asking them both free-response and multiple-choice questions. Complex questions produced up to six times more carbon dioxide emissions than questions with concise answers.
In addition, 'smarter' LLMs with more reasoning abilities produced up to 50 times more carbon emissions than simpler systems to answer the same question, the study reported.
'This shows us the tradeoff between energy consumption and the accuracy of model performance,' said Maximilian Dauner, a doctoral student at Hochschule München University of Applied Sciences and first author of the Frontiers in Communication study published Wednesday.
Typically, these smarter, more energy intensive LLMs have tens of billions more parameters — the biases used for processing token IDs — than smaller, more concise models.
'You can think of it like a neural network in the brain. The more neuron connections, the more thinking you can do to answer a question,' Dauner said.
What you can do to reduce your carbon footprint
Complex questions require more energy in part because of the lengthy explanations many AI models are trained to provide, Dauner said. If you ask an AI chatbot to solve an algebra question for you, it may take you through the steps it took to find the answer, he said.
'AI expends a lot of energy being polite, especially if the user is polite, saying 'please' and 'thank you,'' Dauner explained. 'But this just makes their responses even longer, expending more energy to generate each word.'
For this reason, Dauner suggests users be more straightforward when communicating with AI models. Specify the length of the answer you want and limit it to one or two sentences, or say you don't need an explanation at all.
Most important, Dauner's study highlights that not all AI models are created equally, said Sasha Luccioni, the climate lead at AI company Hugging Face, in an email. Users looking to reduce their carbon footprint can be more intentional about which model they chose for which task.
'Task-specific models are often much smaller and more efficient, and just as good at any context-specific task,' Luccioni explained.
If you are a software engineer who solves complex coding problems every day, an AI model suited for coding may be necessary. But for the average high school student who wants help with homework, relying on powerful AI tools is like using a nuclear-powered digital calculator.
Even within the same AI company, different model offerings can vary in their reasoning power, so research what capabilities best suit your needs, Dauner said.
When possible, Luccioni recommends going back to basic sources — online encyclopedias and phone calculators — to accomplish simple tasks.
Why it's hard to measure AI's environmental impact
Putting a number on the environmental impact of AI has proved challenging.
The study noted that energy consumption can vary based on the user's proximity to local energy grids and the hardware used to run AI models.That's partly why the researchers chose to represent carbon emissions within a range, Dauner said.
Furthermore, many AI companies don't share information about their energy consumption — or details like server size or optimization techniques that could help researchers estimate energy consumption, said Shaolei Ren, an associate professor of electrical and computer engineering at the University of California, Riverside who studies AI's water consumption.
'You can't really say AI consumes this much energy or water on average — that's just not meaningful. We need to look at each individual model and then (examine what it uses) for each task,' Ren said.
One way AI companies could be more transparent is by disclosing the amount of carbon emissions associated with each prompt, Dauner suggested.
'Generally, if people were more informed about the average (environmental) cost of generating a response, people would maybe start thinking, 'Is it really necessary to turn myself into an action figure just because I'm bored?' Or 'do I have to tell ChatGPT jokes because I have nothing to do?'' Dauner said.
Additionally, as more companies push to add generative AI tools to their systems, people may not have much choice how or when they use the technology, Luccioni said.
'We don't need generative AI in web search. Nobody asked for AI chatbots in (messaging apps) or on social media,' Luccioni said. 'This race to stuff them into every single existing technology is truly infuriating, since it comes with real consequences to our planet.'
With less available information about AI's resource usage, consumers have less choice, Ren said, adding that regulatory pressures for more transparency are unlikely to the United States anytime soon. Instead, the best hope for more energy-efficient AI may lie in the cost efficacy of using less energy.
'Overall, I'm still positive about (the future). There are many software engineers working hard to improve resource efficiency,' Ren said. 'Other industries consume a lot of energy too, but it's not a reason to suggest AI's environmental impact is not a problem. We should definitely pay attention.'
Hashtags

Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles


Globe and Mail
20 minutes ago
- Globe and Mail
Prediction: This Unstoppable Artificial Intelligence (AI) Stock Will Join Nvidia, Microsoft, Apple, Amazon, and Alphabet in the $2 Trillion Club by Year's End
Key Points The $2 trillion club is full of businesses benefitting from the growing demand for artificial intelligence. The company I'm eyeing is developing its own AI capabilities that serve multiple cases across its business with huge revenue opportunities. The stock trades for a fair value, and even slight outperformance could push it into $2 trillion territory. 10 stocks we like better than Meta Platforms › Nvidia recently became the first ever $4 trillion company in the world. Its rapid ascension in value stems from growing demand for artificial intelligence. But Nvidia isn't the only company that's seen its market value soar to multitrillion-dollar levels on the back of AI-fueled growth. The three biggest cloud computing providers -- Amazon, Microsoft, and Alphabet -- all boast market caps above $2 trillion. Meanwhile, Apple remains one of the most valuable companies in the world as it works to catch up on its AI capabilities. Where to invest $1,000 right now? Our analyst team just revealed what they believe are the 10 best stocks to buy right now. Learn More » But the $2 trillion club may be about to get a little bigger. One company is showing strong financial results stemming from the rapid advancements of artificial intelligence over the last few years. In fact, I predict it will surpass the $2 trillion market cap milestone before the end of the year. Here's the AI giant that could join the $2 trillion club. One of the biggest beneficiaries of generative AI capabilities I predict that the next member of the $2 trillion club will be Meta Platforms (NASDAQ: META). Not only does it already have a market cap of roughly $1.8 trillion as of this writing on July 24 -- which puts it about 11% from $2 trillion -- but the stock currently looks undervalued relative to the potential opportunities. AI could boost its revenue in the near term while opening up even bigger opportunities in the long run. During Meta's first-quarter earnings call on April 30, CEO Mark Zuckerberg laid out five major opportunities for the company with AI. Improved advertising: Meta has long used machine learning algorithms to help surface advertisements amid organic content to drive maximum engagement. That's led to steady improvements in ad pricing for the company. It's also rolled out generative AI tools that help marketers come up with creatives (ads). In the pipeline, Meta's developing an AI agent that can take a marketer's objective and budget and create and run the entire campaign for them. That has the potential to save marketers money and increase the total number of companies running ads on Meta's properties, further pushing ad prices higher. More engaging experiences: Zuckerberg details two benefits of AI: better recommendations and new types of content. Meta has expanded its AI model to include more data points across all different types of content to improve recommendations across every surface of its apps, including Facebook, Instagram, and WhatsApp. As it grows the model bigger and bigger, it's getting better and better at engaging users. That's only possible because it now has the compute power to support its large language model development. Zuckerberg also expects generative AI tools to provide new ways for creators to produce better content for users. Everything from existing content like photos and videos can be manipulated with AI, and generative AI could enable creators to produce more interactive content as well. Business messaging: Meta's WhatsApp for Business is a relatively small source of income right now. But as Meta improves its AI agent capabilities, it reduces the cost for businesses to provide customer service and sales through WhatsApp and Messenger. That could lead to a surge in WhatsApp for Business users. One analyst thinks AI agents alone are a $100 billion opportunity for Meta. A stand-alone AI chatbot: Meta has integrated the Meta AI assistant into all of its main apps and released a stand-alone version of the app as well. As the user base grows, it could provide another source of valuable advertising inventory. Importantly, since Meta is developing its own large language model for the above applications already, the additional cost of building and running a stand-alone AI chatbot is far lower than for dedicated AI companies like OpenAI or Anthropic. Devices: Zuckerberg points out the growing popularity of Meta's AI glasses. Unit sales tripled in the first quarter. Longer term, generative AI may be essential for creating an augmented reality user interface that fits into the unique setting of each user. Indeed, AI has the potential to dramatically impact Meta's financials in a positive direction in the near term while supporting its long-term objectives in virtual and augmented reality. The stock looks like a bargain right now The above factors should be able to generate strong double-digit revenue growth for Meta for years to come. The company saw 16% revenue growth last quarter, while exhibiting nice operating leverage. As a result, operating income climbed 27% year over year. The big step up in capital expenditures could weigh on earnings growth for the next couple of years as depreciation expense climbs as a result. But as the company grows into those expenses, it should continue to show operating leverage. Meta's also using excess cash flow to repurchase shares. It bought back $13.4 billion worth of its stock in the first quarter, and it still has $70 billion in cash on the balance sheet. As a result, the company should be able to generate strong earnings-per-share growth. As of this writing, the stock trades for 28 times earnings. Considering the growth potential ahead for the stock, that's an enticing price for investors. To push the stock to $2 trillion, it would have to trade for closer to 31 times earnings, which isn't an unreasonable multiple for the stock. But if Meta ends up outperforming expectations, it could trade for the same multiple and still achieve a $2 trillion valuation. I expect a combination of multiple expansion and outperformance to drive the stock to $2 trillion before the end of the year. Should you invest $1,000 in Meta Platforms right now? Before you buy stock in Meta Platforms, consider this: The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Meta Platforms wasn't one of them. The 10 stocks that made the cut could produce monster returns in the coming years. Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you'd have $636,628!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you'd have $1,063,471!* Now, it's worth noting Stock Advisor's total average return is 1,041% — a market-crushing outperformance compared to 183% for the S&P 500. Don't miss out on the latest top 10 list, available when you join Stock Advisor. See the 10 stocks » *Stock Advisor returns as of July 21, 2025


Globe and Mail
10 hours ago
- Globe and Mail
What Are the 3 Best Bargain Artificial Intelligence (AI) Stocks to Buy Right Now
Key Points Taiwan Semiconductor's projected growth should translate into a higher premium than it is currently trading at. The market is concerned that AI will disrupt Adobe's business. Google Search is a primary target of several generative AI companies. 10 stocks we like better than Alphabet › Finding bargains in the artificial intelligence (AI) investing world isn't easy, but they're out there. Three that I've got my eye on are Taiwan Semiconductor (NYSE: TSM), Adobe (NASDAQ: ADBE), and Alphabet (NASDAQ: GOOG)(NASDAQ: GOOGL). All three of these trade at a hefty discount to the broader market, yet are still quite promising. If you're looking for value in the AI space, this trio is an excellent starting point. Where to invest $1,000 right now? Our analyst team just revealed what they believe are the 10 best stocks to buy right now. Continue » How do these three fit into AI? Taiwan Semiconductor is the best-positioned of this trio. It's the primary chip fabricator for some of the leading tech companies, including Nvidia and Apple. Its chip production facilities fabricate chips that its customers have designed, placing TSMC into a neutral position in the AI race. It performed phenomenally well over the past few years, with revenue in U.S. dollars rising an astounding 44% in the second quarter, which exceeded expectations. This strength is expected to persist for many years. Management guided at the start of 2025 that it expects its revenue to increase at nearly a 20% compound annual growth rate over the next five years. Taiwan Semi is a key player in AI technology, and it holds an enviable position. Adobe makes leading graphics design tools that are the industry standard. Whether it's video editing or image creation, Adobe is a top option. However, investors are growing increasingly concerned that generative AI creation technologies could displace Adobe. Image and video creation using generative AI models has come a long way, and has reached a point where it's nearly indistinguishable from what humans create. As a result, many are forecasting the downfall of Adobe. However, I think that's a bit premature. Adobe has also invested heavily in generative AI and has its own Firefly product that allows seamless integration of AI with its existing editing tools. This allows Adobe to compete in this realm while also giving creators more control over the end product than what many generative AI models do. This could keep Adobe relevant within the graphic design industry, making the forecast of its downfall somewhat inaccurate. Currently, Adobe is performing well and has posted consistent revenue growth over the past few years. ADBE Operating Revenue (Quarterly YoY Growth) data by YCharts. If Adobe is supposed to be getting displaced, don't tell it that, as it's still growing at a healthy pace. Last is Alphabet, the parent company of the Google Search engine. Its stock is running into the same fears as Adobe's, as investors worry that generative AI will replace Google Search. While some have made the switch, investors need to remember that Google is an ingrained habit among internet users around the globe. It would take a massive technological leap, which most users won't need anyway, to get them to switch. Google has already implemented the popular AI search overviews feature, which seamlessly integrates search results and AI, and that could be enough to maintain the vast majority of its dominant market share. If Google Search can maintain most of its market share, the stock is poised to move higher, as it trades at a significant discount to the broader market. How cheap is this trio? Alphabet's stock trades at a deep discount to the broader market, despite strong results. GOOGL PE Ratio (Forward) data by YCharts. Considering that the S&P 500 trades for 23.8 times forward earnings, this seems like a reasonable price to pay for the upside that Alphabet provides. Adobe is similarly cheap, trading for 18 times forward earnings. ADBE PE Ratio (Forward) data by YCharts. Taiwan Semiconductor is actually more expensive than the broader market at 25 times forward earnings. However, it's expected to grow at essentially double the pace of the market over the next five years, so this slight premium to the market seems a bit low. As a result, I'm confident labeling Taiwan Semiconductor as a bargain buy right now. Should you invest $1,000 in Alphabet right now? Before you buy stock in Alphabet, consider this: The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Alphabet wasn't one of them. The 10 stocks that made the cut could produce monster returns in the coming years. Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you'd have $636,628!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you'd have $1,063,471!* Now, it's worth noting Stock Advisor's total average return is 1,041% — a market-crushing outperformance compared to 183% for the S&P 500. Don't miss out on the latest top 10 list, available when you join Stock Advisor. See the 10 stocks » *Stock Advisor returns as of July 21, 2025


Globe and Mail
10 hours ago
- Globe and Mail
Is QuantumScape Stock a Buy Now?
Key Points QuantumScape's solid-state lithium-metal battery technology aims to achieve higher energy density, faster charging, and improved safety compared to traditional lithium-ion batteries. The company has partnered with PowerCo, Volkswagen Group's battery manufacturing arm, to produce its technology and support a capital-light business model. Analysts predict that meaningful revenue won't materialize until 2027, with a surge forecast for 2028. 10 stocks we like better than QuantumScape › QuantumScape (NYSE: QS) is leading the charge to revolutionize energy storage with its innovative solid-state lithium-metal battery technology for electric vehicles and other applications. Its cutting-edge battery technology promises higher energy density and faster charging times, while also significantly improving safety compared to traditional lithium-ion batteries. The company has made headlines recently as its battery technology progresses, and the stock has surged 265% from its 52-week low. Here's what investors need to know about QuantumScape's investment prospects and what to expect from the company over the next several years. QuantumScape's battery technology ambitions Last year, QuantumScape partnered with PowerCo, the battery manufacturing arm of the Volkswagen Group. Volkswagen has been a significant investor in QuantumScape since 2012, committing around $380 million, and is the company's largest shareholder. QuantumScape's collaboration with PowerCo looks to advance its QSE-5 technology platform, with the goal of scaling up production at PowerCo's facilities. PowerCo is the "anchor customer" in QuantumScape's expanding technology ecosystem, as the company aims to develop a capital-light licensing model and expand its partnerships with customers worldwide. Instead of incurring heavy manufacturing investments, QuantumScape would license its technology to these partners, reducing its exposure to manufacturing risks and capital demands. QuantumScape is focused on broadening its range of potential licensing agreements with other automotive manufacturers. It also aims to strengthen its relationships with technology partners and key players across the battery value chain. This recent news sent the stock soaring One of QuantumScape's primary goals coming into 2025 was to bring the Cobra separator process into baseline production. Cobra represents a "step-change innovation in ceramics processing" designed to enable a significant improvement in separator productivity. This new process offers a 25 times improvement in heat treatment speed and requires only a fraction of the physical space per film start (compared to its prior-generation Raptor process). In June, the company announced that its Cobra separator process had reached baseline production, a key step toward commercializing its batteries. The successful baselining of Cobra is crucial for enabling higher-volume B1 sample production of their QSE-5 cells, which will be used in a launch program designed to showcase them, with field testing scheduled to commence in 2026. Since the announcement, the stock has surged 190%. Be mindful of the timeline and its finances One thing investors should be aware of regarding QuantumScape is that it is a company in its early growth stage. The company has not generated any revenue to date and continues to incur losses as it spends heavily on research and development. QuantumScape posted an operating loss of $123.6 million in the first quarter and expects to incur significant expenses and ongoing losses into the future. Capital expenditures in the first quarter of 2025 were $5.8 million, primarily supporting equipment and facilities for higher-volume QSE-5 B1 sample production using the Cobra separator process. The company's full-year capital expenditure guidance for 2025 is between $45 million and $75 million. QuantumScape reported $860.3 million in cash and equivalents at the end of the first quarter and anticipates this cash runway is enough to extend into 2028. This runway has expanded since it announced its licensing agreement with PowerCo, which is expected to lower costs and capital requirements compared to previous joint venture arrangements. Is QuantumScape a buy today? I'm excited about the potential of QuantumScape's groundbreaking battery technology, which has the potential to transform energy storage in the future. However, the company is still in the early stages of development and testing, and will continue to incur losses as it brings its vision to life. Analysts have noted that we should not expect meaningful revenue from QuantumScape until 2027, with projections of $51.6 million that year, followed by a significant leap to $265 million in 2028. Until then, we're looking at a story-driven stock that will attract attention but may not offer as much upside from here, especially given its recent surge. Should you invest $1,000 in QuantumScape right now? Before you buy stock in QuantumScape, consider this: The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and QuantumScape wasn't one of them. The 10 stocks that made the cut could produce monster returns in the coming years. Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you'd have $636,628!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you'd have $1,063,471!* Now, it's worth noting Stock Advisor's total average return is 1,041% — a market-crushing outperformance compared to 183% for the S&P 500. Don't miss out on the latest top 10 list, available when you join Stock Advisor. See the 10 stocks » *Stock Advisor returns as of July 21, 2025