Sustaining Intellect in the AI Economy

On May 22, 2026, Microsoft revealed that artificial intelligence (AI) tech is more expensive than paying human employees, and other companies are reconsidering their short- and long-term investments in AI integration. How do we move forward from the economic imperatives created by the AI economy? How do we begin to take back and mitigate the lasting effects that the fast-paced exigence of AI saturation continues to have on our social choices and mental attention? How do we sustain not just the planet, but our intelligence and humanity? Answering this requires shifting our debate away from whether to use AI-powered tools toward how we can use them with intention, consideration, and purpose, and simultaneously demanding sustainable engineering from tech leaders and governing bodies alike.

My argument is not about whether we should use large language models and other AI-powered tools, but rather that we should use AI with more purpose, more consideration, and with more intention, while demanding more sustainable engineering from tech leaders and governing bodies.

How to sustain our complexities

In conversations about sustainable AI, we often focus on concerns around electricity, water waste, and the environmental impact data centers have on their neighboring communities. Obviously, we should center these material effects in sustainability conversations because the costs are real, going beyond financial waste and into vulnerable parts of society.

However, after using AI in the classroom and gathering student opinions on the state of sustainability, I have become just as interested in how AI has unseen effects on thinking, learning, and understanding. Like most technological advances, the dawn of AI promised a utopian transformation that would wipe away the mundane labor of daily communications, data crunching, and a host of other repetitive tasks in return for more time and space to create. Who wouldn’t want to be freed from the long haul of emails, memos, and presentations and have them look finished before anyone has actually done any thinking? Critical thinking is quite challenging, and most of the time, sloppy AI output shifts the burden of judgment from the sender to the receiver. At the BlackRock Infrastructure Summit on March 11, 2026, OpenAI CEO Sam Altman framed artificial intelligence as a cheap, ubiquitous utility, stating, “We see a future where intelligence is a utility, like electricity or water, and people buy it from us on a meter.” Yet, just eleven days later, Microsoft stated that AI is actually more expensive than paying human employees. This stark contradiction between two industry giants perfectly illustrates the financial and economic dissonance surrounding the realities of AI, while revealing a dystopian push to outsource human intellect and reduce the hard-earned uniqueness of learning into a cheapened, metered commodity.

The saturation of AI marketing makes it intentionally difficult to avoid AI usage for daily professional and personal tasks. My argument is not about whether we should use large language models (LLM) and other AI-powered tools, but rather that we should use AI with more purpose, more consideration, and with more intention, while demanding more sustainable engineering from tech leaders and governing bodies. Sustainable AI should mean cleaner data centers and more efficient models for public use. As a communications instructor, I would argue that sustainable AI should also mean communication practices that sustain our human attention, careers, trust, learning, cultural nuances, and critical judgment. The better question for business leaders is not, “Can AI do this?” but, “Is this work worth automating?” Where should human judgment be sustained and even monetized in this rocky AI economy, and how should education incorporate this human-AI collaboration in a way that prizes actual human thought and the creative value of our thinking processes?

Saving context in the classroom

I want to reflect on my own encounters with the issue of sustaining human context in the classroom. AI is included in the learning outcomes of the courses that I teach, and the most useful question was, “What kind of learning does AI make possible without shortcutting cognitive development?” I did not want AI to replace the classroom content, grade for me, generate assignment feedback, or flatten student voices into the same prose that it often produces. At the same time, it is difficult to track AI usage in assignment output since most people report inaccuracies with AI detection tools. I wanted to know whether it could help students practice revising a message for a different audience, anticipating questions for skeptical stakeholders, comparing tones across cultures on digital platforms, or moving from a vague claim to a more specific recommendation. The more I interacted with AI, the more I found that it often excludes or truncates deeper meanings and does not capture the depth of context in the communication being produced. Georgia Tech PhD student Xinjie Shen wrote, “LLMs’ tendency to directly answer when crucial context is missing and their significant performance degradation as a result … leads to completely wrong answers.” Capturing cultural context, for example, is often more challenging due to the human peculiarities of language and social behavior. Each business or organization has a culture that cannot be outsourced, electronically transmitted, or excluded from human attention and care.

During October 2024, my colleague Dr. Rima Bhattacharyay and I received a grant from the Center for Intelligent Business to create a voice-to-voice AI application aimed at enhancing students’ intercultural conversational skills and their understanding of cultural nuances in a variety of international business contexts. We were tasked with treating the LLM as a collaborative agent and analyzing how AI-powered translation tools handle cultural nuances, idioms, and context-specific meanings. Thinking through human accent, cultural subtleties, common intercultural business values, and vocal interactivity were the hardest parts of building the agent and describing cues to prompt cross-cultural understanding. However, the focus on human relationships across intercultural contexts felt like a keener way to sustain cultural diversity against AI’s tendency to compress discursive context.

In the 2025 academic year, I revised the AI agent, following my interest in earth-honoring traditions to center sustainability communication in business consulting. The new version simulates a realistic executive-client check-in meeting with business communication students who were proposing a sustainability strategy that integrated at least three United Nations Sustainable Development Goals into a company’s internal and external operations. In a follow-up survey, students shared that the conversation was valuable preparation for the subject matter that they needed for their upcoming in-class meeting, and they suggested the AI application be made more authentic by including follow-up questions and natural interruptions that occur in real meetings. Some wanted more conversational nuance and better context from the agent.

By the end of the project, students discovered a wealth of business solutions and became critical of how businesses communicated and enacted the environmental, social, and governance (ESG) policies for employees and external stakeholders. Communicating sustainability needs and clearly articulating material demands are practical ways to guide future AI expansion toward a more sustainable and collaborative human-nature-technology ecology.

Who to trust now?

As a point of class discussion, I asked students what they thought they could do right now and early in their careers to manifest a more sustainable reality. While most students believed their everyday choices mattered, one countered that individual agency is an illusion when large corporations dictate the market of sustainability. From its last tally in 2025, only about 18 percent of the UN’s 2030 Sustainable Development Goals have been achieved, so it is reasonable to harbor corporate skepticism in a culture that does not prioritize sustainable data centers and legal parameters around AI job replaceability. How can we trust companies to abide by the triple bottom line if they only value profit over human cognition and education?

In industries that measure success with revenue, productivity, and efficiency, sustainable AI requires a unique kind of discipline: a willingness not to automate everything that can be automated and a commitment to centering true human learning and context over machine-driven business models. Developing critical thinking skills is challenging, so while I understand the perceived efficiency requirement to outsource our thinking, we should protect the friction of critical thought and human context at all costs. This friction is where learning, ethics, cultural awareness, and judgment are properly aligned. The companies that understand this will not ask only how much AI can produce. They will ask what kinds of human capacity their AI usage is building or collapsing.