Emily Niebuhr is an English junior and an opinion columnist for Mustang News. The opinions expressed in this article do not necessarily reflect those of Mustang Media Group.
You’ve got a date tonight, and this time, he’s different. He’s reliable, he compliments you and he agrees with you about nearly everything. He speaks intelligently, eloquently weaving together threads of information from every corner of the internet.
His name is ChatGPT, and Cal Poly personally introduced you two. But who’s really behind the seductive glow of your laptop screen? That’s Sam Altman, OpenAI’s CEO, and someone just threw a Molotov cocktail through his window. Evidently, not everyone’s relationship with AI is going so well.
In February 2025, the Cal State system signed a $17 million contract with OpenAI that granted Cal State students, faculty and staff free access to ChatGPT Edu through July 2026. This AI initiative has been met with mixed reactions. Aside from ethical concerns with cheating, it raises a watershed of environmental and philosophical dilemmas.
READ MORE: Cal State system AI initiative sparks faculty concerns regarding privacy
We need to ask ourselves how heavily we should rely on AI and consider how it’s impacting our ecosystems. It’s a sensitive topic because so many of us have mindlessly folded AI into our study routines. The uncomfortable reality is that every seemingly inconsequential conversation with ChatGPT increases the demand for energy and water, resources we are already straining, that could be allocated elsewhere.
We are also guinea pigs for tech companies whose end goal is to create dependence on their models. This may sound extreme, but Altman said, “We see a future where intelligence is a utility like electricity or water and people buy it from us on a meter,” according to previous coverage by Kevin Okemwa at the March BlackRock Infrastructure Summit.
This comparison puts a price on intelligence and implies the same dependency we have on water.
Does ChatGPT really use that much water?
Altman dismissed water use from data centers by comparing AI to humans in an interview with the Indian Express at the India-AI Summit 2026, claiming “But it takes a lot of energy to train a human. It takes like 20 years of life and all of the food you eat during that time before you get smart.”
This comparison is a false analogy, unless Altman is seriously proposing the replacement of human intelligence with AI, which would be very concerning. Humans consume energy and resources out of necessity, and AI consumes them extraneously. We need food; we don’t need to ask ChatGPT what the answer to that math problem is.
Water is not the infinite resource we’d like to imagine it as. Only 3% of Earth’s water is fresh, and less than 1% of that water is accessible for human use. On top of that, we are depleting freshwater and groundwater resources faster than the increasingly unstable weather patterns can replenish them.
There are currently 4,088 data centers in the United States, according to the USA Data Centers Map and the Open Source Data Center Atlas, and estimates for daily water consumption vary widely, with some data showing a single center can consume up to 5 million gallons per day. Alongside water use, casual queries directly contribute to significant energy use, with a single AI-focused data center annually consuming as much energy as 100,000 households.
How does this affect students at Cal Poly?
ChatGPT Edu is a double-edged sword that can sharpen our knowledge, but it can also cut us.
Every interaction with AI degrades our ability to self-sufficiently analyze information and do what Chat cannot: create new ideas.
If students decide to use it, they should learn to recognize its shortcomings. Vidya Schalk, a bioresource and agricultural engineering lecturer, believes students should learn how to work with AI.
“We have to prepare the students to use it in a very careful way and use it as a tool, not to have a cognitive downgrade,” Schalk said.
ChatGPT Edu can quickly turn into an echo chamber, especially if we approach the information it provides as the end rather than the means. Although useful for quick calculations and data synthesis, it shouldn’t be where research starts.
“It’s pattern matching, it’s able to pull all these threads,” Schalk said. “That’s what it’s doing at a massive scale, but it’s not creating anything new.”
Students can be academically successful without any reliance on AI. Ela Sims-Mahoney, a psychology junior, said she rarely uses large language models like ChatGPT to assist her studies.
“I’m not the biggest fan of AI in general,” Sims-Mahoney said. “I think it’s really easily manipulated, and people can become dependent on it really quickly.”
Altman spoke of AI as a source for intelligence, but what ChatGPT offers you isn’t knowledge. We cannot conflate true research with information that has been scraped from every corner of the internet and generalized into a digestible paragraph.
“You can call it privatization, or you can call it democratization,” Schalk said. “It’s a cognitive surrender that we’re going to have. That’s what is going to inhibit our minds.”
“At the end of the day you have a controllable population, more easily exploitable and disposable. But if people are taught how to use it, that it’s a tool, it becomes an aid. It’s like a partner.” Schalk said.
What should a healthy relationship with AI look like?
You don’t have to break up with Chat, or throw a Molotov cocktail through its window. But you should question its intentions as well as the intentions of billionaire tech CEOs like Altman.
Ask yourself: Is free access to ChatGPT Edu a gift of knowledge or an avenue for facilitating dependency? If intelligence is a resource, who’s controlling it?
If we don’t carefully consider these angles before engaging with any form of AI, data centers won’t just create water scarcity, but also a shortage of the literacy, problem-solving skills, and critical thinking necessary to address today’s most urgent issues, like climate breakdown.
Don’t use AI casually and excessively. There are wide-reaching effects for your mind and the communities built near these data centers. When you do use it, turn to it as a tool or last resort, not the starting point.