Introduction
Grok 4 marks a breakthrough in artificial intelligence, but its training demanded unprecedented energy and financial resources. Discover the key numbers and real impact of this technology.
Context
The AI sector is rapidly evolving, and models like Grok 4 push innovation boundaries while raising environmental and cost challenges.
Grok 4’s Energy Footprint
Training Grok 4 required enough electricity to power a town of 4,000 people for a year. Most of this energy came from natural gas, generating emissions equal to a Boeing jet over three years.
Direct definition
Grok 4 demanded exceptional energy, with significant emissions and a tangible environmental impact.
Costs and Infrastructure
The supercomputer used for Grok 4 needed cooling equivalent to 300 Olympic-sized pools. Energy and chip costs alone reached $500 million, not including expenses for serving the model, research, and human labor.
- Energy: enough for a small town
- Emissions: similar to a commercial jet
- Cooling: 300 Olympic pools
- Cost: $500 million for training alone
Risks and Limitations
These figures highlight the environmental and financial risks of advanced AI. The impact extends beyond training, as serving and research continue to generate costs and consumption.
Conclusion
Grok 4 shows that AI innovation brings real challenges in sustainability and investment. Transparency about costs and environmental impact is crucial for the sector’s future.
FAQ
What is Grok 4’s energy footprint?
Grok 4 uses enough energy to power a town of 4,000 people for a year.
How much does training Grok 4 cost?
Training alone required $500 million.
What are the main energy sources used?
Most energy comes from natural gas.
What is Grok 4’s environmental impact?
Emissions are comparable to a Boeing jet over three years.
How much cooling is needed?
The supercomputer needed cooling equal to 300 Olympic pools.
Are there other costs beyond training?
Yes, serving, research, and human labor add significant costs.
Why is it important to know these numbers?
It helps assess the sustainability and real impact of advanced AI.