- This topic is empty.
-
Posts
-
AI Wants More Data. More Chips.
More Real Estate. More Power. More Water. More Everything ☹.
At first glance, interacting with ChatGPT seems simple: ask a question, and it provides an answer. But behind the scenes, every ChatGPT prompt and Microsoft Copilot task involves massive resource consumption. It takes millions of people working to engineer, correct, and train the models. The energy used could power entire countries, with data center mega campuses spanning the globe. It relies on power lines, internet cables, water, land, and essential metals and minerals. Artificial intelligence requires all of this—and even more in the future.Research suggests that a single ChatGPT query consumes nearly ten times the electricity of a traditional Google search. While a typical search engine pulls information from a vast, pre-indexed web, modern AI products like ChatGPT use large language models (LLMs). These models are trained on billions of words—from Shakespeare’s works to the latest Federal Reserve forecasts—identifying patterns and creating billions of parameters to simulate human behavior. This is how generative AI, like ChatGPT, creates new content.
The resource demands of AI will inevitably shape the industry, creating a divide between those with abundant resources and those without. Access to the necessary materials, like chips and rare minerals, will become increasingly competitive. To stay ahead, tech companies are investing billions into alternative energy solutions like nuclear fusion, hoping to overcome decades of stagnation. However, the ever-growing need for energy is also pushing the world to rely more on fossil fuels to sustain power grids, even as global emissions targets slip further out of reach.
While the AI revolution offers immense opportunities for investors, businesses, and societies, it’s not without risks. Concerns about potential biases and harms from AI systems continue to surface, while Wall Street grows impatient for tangible profits. The focus on efficiency may also backfire for those who over-invest in infrastructure.
Keep reading below for an in-depth look at what the
AI industry requires to maintain its ever-expanding operations.
Such as, Another 1,000 Terawatt-Hours of Power, All the Fossil Fuels It Can Get—and More, A Hundred Times More Grid Capacity, Billions of Liters of Water a Day, Twice as Much Internet Bandwidth, Real Estate for Thousands of Data Centers, Chips, Chips, Chips, Silicon, Steel, Quartz and Copper, More People Than You Think, More (Good) Data Than the World May Have, and Maybe Less of Everything Than Some Fear. Or Hope.
- You must be logged in to reply to this topic.