On Sept. 8, Jacob Coxon, a pretraining researcher at Anthropic, announced his resignation from the company via his X account. Prior to this, he had worked for both OpenAI and Anthropic for three years, citing irresponsible actions by both organizations for his abrupt departure.
CEO of Anthropic, Dario Amodei, speaks to UN members remotely, during an AI Security Council meeting on Wednesday, Sept. 23.
| Yuki Iwamura, AP News |
His statement reached millions of viewers overnight, including Dr. Gary Rose, a Political Science Scholar-in-Residence at Sacred Heart University.
“Listening and watching Coxon on television was truly disturbing, particularly because of his expertise related to AI,” said Rose. “What he was warning about needs to be taken seriously.”
In his X post, Coxon’s main warning states that AI companies “are racing to self-improving superintelligence and gambling with our lives.” He goes on to stress how powerful this technology can be, and without anything slowing them down, the “superhuman systems” will be able to “hack anything, revolutionize any field overnight, and acquire real power and resources.”
Talks about regulations on AI development have become increasingly popular. To Professor Robert McCloud, Director of Technology Integration, checks on AI are a no brainer.
“You can’t put people on a jet that hasn’t been tested and approved,” said McCloud. “You can’t give somebody a drug [that hasn’t been tested]. You can’t even buy hamburger meat that hasn’t gone through testing.”
McCloud emphasized independent regulation as a possible solution.
“You don’t ask the foxes–OpenAI, Meta, Google, ‘how should we regulate you?’” said McCloud. “It has to involve embedded, high salary people at the companies that have the power to put the brakes on and say, ‘you can’t release that.’”
There are mixed opinions with how much regulation is necessary. Rose cautions against too much government oversight, but too little could lead to harmful consequences.
“Government working closely with AI executives to monitor a sustained but not overly excessive development in this new technology is probably the best approach,” said Rose.
Professor Arc Han, from the political science department, described a triangular mechanism to potentially slow down the progression of AI.
“We should have some civil organization, not just government and business, to balance and play a role between companies and government,” said Han.
Some, like junior Gianmarco Pirri, a marketing major with an applied AI minor, question if oversight is even possible.
“The regulation of AI might seem like it should be explored but it would be incredibly difficult to do,” said Pirri. “Theoretically, it would help to prevent issues that we’ve recently read about in media, but it also makes you ask, ‘have we reached a point where it isn’t fully possible to be regulated?’”
However, many AI executives are in favor of regulation. CEO of Anthropic, Dario Amodei, as well as Sam Altman (OpenAI) and Elon Musk (SpaceXAI) have all called for greater government oversight of AI, according to AP News.
However, a concern for Rose lies in the intense polarized state of American politics, making it hard for Congress to pass legislation that will enforce the regulation.
“Presidential leadership, rather than depending on congressional leaders, offers a more direct response in this policy area,” said Rose.
When responding to efforts to regulate AI technology, President Trump posted on Truth Social, “AI taking over the World, destroying Humanity, and all other things bad, is a HOAX.”
Regardless of the circumstances, Rose urges everyone to keep an open mind about AI.
“Listen to what experts on both sides of the argument have to say,” said Rose. “AI is an evolving technology that one way or another will have global consequences.”
As for Han, he is hopeful that the world will work through this, but there are certainly some dangers.
“I do believe that human beings have these agencies, if things are really going badly, people will stay together,” said Han. “Big changes happen many times in history, but we have adapted. The problem is how much cost we’re going to pay and who bears that cost in the process of transition.”
