If AI ends the world, it won’t be because of AI.
It will be because of human greed.
In September 2026, Jacob Coxon, a former OpenAI and Anthropic AI researcher, quit his job at Anthropic over concerns about self-improving AI models. His resignation announcement on X prompted other researchers to disclose that they believe AI could cause human extinction, some predicting by 2030, unleashing a wave of concern about AI development.
But the danger doesn’t just lie in the AI race itself. It’s also in the human greed that started it and the relentless desire that continues to fuel its vicious cycle, despite the risks.
UCLA is no stranger to collaborating with companies that power AI. In 2026, the UCLA Henry Samueli School of Engineering and Applied Science launched a $125 million semiconductor hub in partnership with Applied Materials, Broadcom, GlobalFoundries, Meta and Synopsys.
[Related: UCLA announces $125 million Semiconductor Hub with industry partners]
Labs in the Department of Computer Science also partner with companies such as Amazon and Coco Robotics to conduct their research.
At a time when the incentive to build advanced technology is greater than ever, UCLA research labs partnered with AI companies need to consider the greater benefits and harms of their work.
There is corporate competition between AI companies to build the most powerful and superintelligent model. With about half of U.S. adults using AI, it’s no surprise there is both a financial advantage and cultural prestige granted to the top AI systems.
In 2025, ChatGPT was the most downloaded app globally, with its creator, OpenAI, projected to earn $50 billion in revenue in 2026, according to Bloomberg.
Max Mendoza, a third-year mathematics and philosophy student and president of Bruin AI Safety, said he believes there are currently no socioeconomic incentives to encode safety in AI models.
“Currently, the primary bottleneck is shaping the incentives of various actors, whether these are individuals, companies or countries,” Mendoza said.
This race has expanded globally. The AI arms race between China and the U.S. adds another layer of competition that fuels the drive to build even more.
But the competition did not arise on its own. The ambitions of tech leaders, politicians and investors have fueled a race in which the desire for financial and industrial dominance makes falling behind unacceptable. With every new model, the stakes grow, igniting competing companies to work harder.
What’s most concerning is that this competition makes it easy for companies to disregard the imminent risk researchers like Coxon warn of.
For example, in 2023, OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei and Google DeepMind CEO Demis Hassabis signed a warning stating that mitigating extinction due to AI should be treated with the same seriousness as pandemics or nuclear war.
Despite this, all three companies are continuing to develop powerful AI models.
William Chang, the founder of BruinML – a machine learning group at UCLA – said he is skeptical of whether superintelligence is in our near future.
Chang, a doctoral student in applied mathematics, added that progress isn’t linear. While AI models can solve leading mathematical problems, they still cannot show empathy and other capacities that come naturally to humans.
However, even if superintelligence isn’t possible by 2030, the competitive pressure to build the world’s greatest AI may eventually lead us there.
“If OpenAI isn’t developing the new insane chatbot, somebody else will,” Chang said.
When slowing down development can be seen as losing, UCLA researchers collaborating with AI companies must consider the consequences of their work, even when doing so challenges our personal ambitions.
[Related: UCLA’s digital humanities major must counter AI bias, shape technology responsibly]
Certainly, with innovation comes the discovery of new consequences. But, considering the warning signs, it is a responsibility Bruin AI developers have as researchers and innovators.
“The field needs people who can both build these systems and question them,” said Lucy Avetisyan, co-chair of the UCLA AI Committee, in an emailed statement. The committee guides the university’s responsible, ethical and innovative use of AI.
It is imperative students and faculty building AI systems in partnership with AI companies do not let the pursuit of technological advancement blind them to the risks progress may pose. As a leader in research, UCLA holds significant power to shape the future of technology. It’s important that we don’t fall victim to unchecked ambition.
If we do, perhaps Coxon is right.
Comments are closed.