With the European Union, G7 members, and China moving faster than other countries despite many questions still swirling, AI leaders like Ahmed Reza are encouraging governments to work together with the industry.
Reza is a prominent AI expert with decades of experience in the field of AI technology, having used it since his early days as a student at Cornell University. Self-described as an ‘AI Geek’, Reza is the founder of AI startup Yobi and one of the founders of the entrepreneurial community TrepHub. Having helped hundreds of tech entrepreneurs make their ideas a reality just like he did, Reza is now calling for increased collaboration among Middle East and AI industry leaders.
“Leaders all over the world are already aware of the importance of getting this right,” says Reza, who has previously commented on the American and European approaches to AI regulation. “Unfortunately, there is a very thin line between safeguarding ethical standards and promoting innovation when it comes to disruptive technologies.”
The potential uncertainties associated with AI, including those related to job displacement, fraud, cybersecurity, and disinformation, largely account for the hasty regulatory responses. However, the AI community in general shares Reza’s concerns about regulators going too far due to a misunderstanding of the technology and an overeagerness to act.
OpenAI’s Sam Altman has openly stated that concerns about human obsolescence are overestimated, while also hinting his firm could leave the region if the regulations were too harsh. While Altman has explained in the past that models like ChatGPT are ‘incredibly limited’ in reality and it is a ‘mistake to be relying on it for anything important’, this hasn’t prevented people from using it in situations as important as legal proceedings.
“The media, the public, and lawmakers are being deceived by these models’ ability to generate convincingly human-like responses,” explains Reza. “Unless an AI model is really specialized in a specific topic or use case, it is very easy for them to hallucinate and provide false information. It’s crucial to recognise the limitations of these models and exercise caution when relying on their outputs for important decisions or assessments.”
Throughout history, lawmakers have often had a difficult time understanding the technical aspects of new technologies, as well as their implications. Phrases like ‘Senator, we sell ads‘ and ‘A series of tubes‘ have now become synonymous with the challenges this lack of expertise means when it comes to technology and legislation.
“It is completely understandable that regulators don’t understand how every new technology works but the problem comes when some of them refuse to collaborate with those who do. Fortunately, many leaders are open to having these difficult, yet essential, conversations to take actions that facilitate innovation while safeguarding society,” explains Reza, referring to the virtual conference he will have with leaders of the Middle East on May 30.
As AI continues to advance and find applications in various domains, it is crucial to address its challenges proactively. Reza’s emphasis on recognising the limitations of AI models and exercising caution when relying on their outputs for important decisions or assessments is a timely reminder. However, only time will tell how effectively governments and the AI community can navigate the complexities of AI regulation.