Op-ed: Take a deep breath: DeepSeek is just the start of the artificial intelligence transformation

By Nathan McNeese

Artificial intelligence holds great promise for making our lives safer and easier, but its rapid development raises questions about whether we can control it and ensure it serves the best interests of humanity.

DeepSeek, an impressive feat of computer engineering, is an excellent example of just how fast AI development is moving.

News of DeepSeek’s emergence stunned Wall Street and underscored that the United States is locked in a high-stakes global AI race with multiple countries. To the average user, DeepSeek is just as effective as comparable chatbots, yet it was created for a fraction of the cost and computing power.

If you are concerned with the potential impacts of AI, you have good reason to be. But I would advise taking a deep breath because we are just getting started.

AI is progressing at a rate unprecedented for technology, faster than almost anyone predicted. Major developments like DeepSeek are likely to keep coming for at least the next decade.

Opinions within the United States about whether the developments are positive or negative will vary. But from an even larger perspective, there will be major variance among nations, leading to global challenges.

Again, DeepSeek offers a glimpse of what could lie ahead. Built on open-source code, DeepSeek allows external developers to modify and expand it. That opens the door for rapid innovation but also raises concerns about misuse by unqualified individuals—or those with nefarious intentions.

The push to win the AI race often puts a myopic focus on technological innovations without enough emphasis on whether the AI has some level of understanding of what is safe and right for human beings.

These rapid developments are bringing us closer to what once seemed science fiction– and the stakes are rising.

As little as two years ago, I would have expected that artificial general intelligence (AGI) would take at least 20-30 years to create. Now, we appear to have narrowed that window to more like five years.

While definitions of AGI vary, I see it as artificial intelligence with close to the same abilities as humans in many ways — not only to reason but also to understand cognition and emotion and the ability to have aspects of consciousness.

When AGI becomes a reality, the potential for society to leverage this technology and to improve and expand will be at an all-time high. Jobs that are not optimal for humans will be entirely replaced with AI, but new professional careers and opportunities will be created.

However, the equal opportunity for society to misuse AI will match this meteoric rise. What determines the path forward is the approach we take over the next decade.

So what can we as human beings do to ensure AI serves the best interests of humanity?

Companies that are developing AI need to look beyond money and do what is right for human nature. Human-centeredness needs to be built into AI models, and those models should be thoroughly tested with human beings before they are released to the masses.

Public education also has an important role to play. We should be teaching students to better understand how AI works and to have a healthy amount of skepticism toward AI systems, which sometimes make mistakes. We cannot get to a place where we are blindly using this technology without ensuring that we as humans are verifying and validating it.

Currently, we are not providing good educational materials and AI user guides to understand this technology. Combine that with how fast it is moving, and we are most likely headed for a point in which this technology will be so advanced that a wide majority of humans will have no idea what they are interacting with– or when, where and how they should be interacting with it.

AI can be an amazingly powerful technology that benefits humanity if used correctly. Yet, as a society, we need to be better at making certain that AI is being used and designed in a manner that is fully working for us in a safe and effective manner, and not the other way around.

In the future, we, as humans, must ensure that this is the paradigm: we are in control and in charge of AI. We need to try to minimize the bad through oversight and education, and we need to maximize the good by figuring out how we, as humans, can utilize AI to help us make our lives better.

There continues to be great promise for AI to change all of our lives positively, but we are in a critical time in history to ensure that things go the right way.

Nathan McNeese is the McQueen Quattlebaum Endowed Associate Professor of Human-Centered Computing and Director of the Center for Human-AI Interaction, Collaboration, and Teaming at Clemson University and a member of the National Academies of Sciences, Engineering and Medicine’s Board on Human-Systems Integration.