Observed Signal · Nov 15, 2022 · Research Study · Source: Trending Topics · Impact: 2/5 · Sentiment: Negative
AI's Carbon Footprint Significantly Higher Than Expected, Says Hugging Face
Hugging Face developed a new method to estimate the full lifecycle CO2 emissions of large language models. Applying it to its own model Bloom, the company found that training produced 25 tonnes of CO2, but including hardware manufacturing and ongoing operation doubles the total to 50 tonnes—roughly equivalent to 60 London-to-New York flights. Bloom's daily use emits about 19 kg of CO2, similar to a 54-mile car trip. External estimates cited in the study put OpenAI's GPT-3 training above 500 tonnes and Meta's OPT above 75 tonnes. The research, reported by MIT Technology Review, notes that the global tech sector accounts for 1.8–3.9% of worldwide CO2 emissions and calls for a better understanding of AI's environmental impact.
Reveals higher-than-expected carbon footprint of LLMs, adding environmental scrutiny to AI and AdTech infrastructure decisions.
Track Hugging Face Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Hugging Face estimated that training its LLM Bloom produced 25 tonnes of CO2.
- Including hardware manufacturing and operational energy doubles Bloom's lifecycle emissions to 50 tonnes of CO2.
- Bloom's daily operation emits approximately 19 kg of CO2, equivalent to a 54-mile drive in an average new car.
- External estimates cited in the study suggest OpenAI's GPT-3 emitted over 500 tonnes of CO2 during training, while Meta's OPT emitted over 75 tonnes.
- The global technology sector accounts for an estimated 1.8% to 3.9% of worldwide CO2 emissions.
Connected Companies & Entities
5 Entities mapped“The US company Hugging Face developed a new approach to measure LLM CO2 emissions and estimated the emissions for its own LLM Bloom....”
“The models GPT-3 from OpenAI and OPT from Meta were estimated to emit more than 500 and 75 tonnes of CO2 during training, respectively....”
“The models GPT-3 from OpenAI and OPT from Meta were estimated to emit more than 500 and 75 tonnes of CO2 during training, respectively....”
“The higher-than-expected climate impact of these models was reported by MIT Technology Review....”
“Many companies such as Google or Meta use AI models to classify user comments or recommend content....”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Xbox Launches TV & Film Division; Meta Tests Link Restrictions
This AdExchanger news roundup covers three major stories. Xbox has unveiled a new TV and film division named XP, led by Kayleen Walters, to explore monetization opportunities including formalizing sponsorships and brand partnerships. Separately, Meta is testing Meta One, a subscription bundle, and has begun charging non-subscribed business pages for including more than two external links in posts, with news pages currently exempt. Additionally, the article discusses the trend of mid-sized independent agencies merging into hybrid holding companies, citing Wpromote's acquisition of Giant Spoon, Chemistry's acquisition of Colossus, and Acadia's purchase of Crush, as competitive pressures from larger groups like Omnicom-IPG and Publicis intensify.
OpenAI Fires Safety Researchers; GPT-6.1 Ultrafast Launch
This AINews newsletter covers several significant developments: OpenAI has fired three safety researchers linked to a METR audit, citing mishandling of confidential information. OpenAI also launched GPT-6.1 Sol Ultrafast, a faster and cheaper model, and introduced an intelligent UI for ChatGPT. Anthropic released Claude Haiku 5.5 and cut Sonnet 5.5 cache-read prices. Additionally, Arena raised $200M, and there are reports about OpenAI's revenue figures. The newsletter also includes various model launches, benchmarks, and safety research updates.
AI Price War: OpenAI Gains Ground on Anthropic
The AI price war is intensifying, with OpenAI gaining significant ground on Anthropic among business customers. According to a Wall Street Journal report, spending on OpenAI and Anthropic models via the OpenRouter platform was nearly evenly split in September among roughly 120,000 companies using both, a shift from January when Anthropic held about 75% of that spending. OpenAI's aggressive price cuts on its GPT-5.6 lineup, including an 80% reduction on its smallest model Luna and 20% on Terra, are driving this change. Companies are increasingly prioritizing cost, combining multiple providers and using cheaper models for simpler tasks. Anthropic faces its own challenges, including capacity issues with Claude Code and data retention criticism. Both companies are preparing for IPOs, needing to demonstrate sustainable revenue to justify valuations exceeding $1 trillion.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.