OpenAI CEO Sam Altman has expressed concerns about the potential risks that come with the development of artificial intelligence (AI) as the company releases its latest model, GPT-4.
Altman, who oversaw the creation of the controversial consumer-focused AI application ChatGPT, warns that the technology could pose a significant danger to society if not regulated and monitored carefully.
Speaking to ABC News, Altman stressed the need for both regulators and society to be involved in the development and implementation of AI technology to safeguard against potentially negative consequences. He also emphasized the need for caution, stating that "we've got to be careful here," and adding that people should be happy that there is some level of fear surrounding AI.
AI Could Also be a Significant Technological Advancement, Says Altman
Altman is particularly concerned about the possibility of large-scale disinformation campaigns and offensive cyber-attacks, which could be facilitated by increasingly sophisticated AI models. Despite these risks, he acknowledges that AI could be one of the most significant technological advancements humanity has ever made.
The release of GPT-4 comes less than four months after the original version became the fastest-growing consumer application in history. While the new version is far from perfect, Altman explains that it has already demonstrated impressive potential. GPT-4 scored 90% on the bar exams in the United States and performed exceptionally well on the high school SAT math test. The model is also capable of writing computer code in most programming languages.
Elon Musk and Altman Share Concerns About AI
The fear surrounding AI often centres on the potential for machines to replace human workers. However, Altman clarifies that AI is a tool that requires human direction and input to function. He acknowledges, though, that some humans may use AI in ways that do not prioritize safety or ethics.
Altman's concerns are echoed by Tesla CEO Elon Musk, one of OpenAI's earliest investors when it was still a non-profit company. Musk has repeatedly warned that AI or artificial general intelligence (AGI) is more dangerous than a nuclear weapon. He has also expressed concern over the lack of regulatory oversight of AI.
AI Is Better as a Tool For Reasoning
Altman warns that GPT-4 and other AI models are not simply a database of facts but rather a reasoning engine that can lead to bizarre responses if not directed properly. He cites the "hallucinations problem" as a potential issue, where the model may confidently state things as facts that are entirely made up. Altman emphasizes that AI needs to be used as a tool for reasoning rather than memorization.
As the development of AI continues at a rapid pace, it is crucial for society to consider the potential risks and benefits. The rise of AI requires ongoing attention and regulation to ensure that the technology is used ethically and safely, and that the benefits outweigh the risks.
Misbar’s Sources:
