Menaka Guruswami writes: Why should AI be regulated?

The areas of concern Altman highlights are important for all countries. A few weeks after Mr. Altman testified before the U.S. Senate, a statement caught my eye. Over 350 people signed it, including Altman. Mira Murati, Chief Technology Officer of OpenAI. Microsoft Chief Technology Officer Kevin Scott. Google AI top executive. This includes Skype and Quora leaders, and even the former UN High Representative for Disarmament Affairs. They all signed a statement stating that “mitigating the risk of AI-induced extinction should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.” ing.

continue, wall street journal AI experts and technical chiefs, including Elon Musk, reportedly signed a letter in March 2023 to “temporarily suspend training systems more powerful than GPT-4.” GPT-4 is a technology released earlier this month by a Microsoft-backed startup. This includes OpenAI’s next-generation technology, GPT-5.

The profitability and efficiency of AI will ease the need for regulation. NVIDIA, the semiconductor company described by The Wall Street Journal as “at the center of the artificial intelligence revolution,” recently became a trillion-dollar company. This puts the company among the elite companies of Apple, Microsoft, Amazon and Alphabet (Google’s parent company). Both are trillion dollar companies. The outlook is bleak as the entire tech industry cuts its workforce. Nvidia provides chips and software to meet the compute-intensive demands of generative AI. ChatGPT is an example of generative AI. As Techopedia.com explains, generative AI can generate a variety of content such as text, images, voice, and synthetic data.

The use of AI has increased significantly over the last few years. Now it’s used to guide weapons, drive cars, administer medical procedures, and even make (often inaccurate) legal notes. In February, the United States launched an effort to foster international cooperation on the responsible use of AI and autonomous weapons by the military. As The Independent reported, this was done in recognition that AI could change the way wars are fought.

The Ukraine war, for example, has accelerated the deployment of AI-powered drones “used to identify, select and attack targets without human assistance.” Ukraine has already deployed semi-autonomous strike drones. Ukrainian Minister for Digital Transformation Mikhail Fedorov said fully autonomous killer drones are the “logical and inevitable next step” in weapons development.

Whether fully autonomous killer drones will be programmed to operate in accordance with the Geneva Conventions prohibiting targeting of civilians and non-combatants is an important concern. At the moment, drones require humans to select targets over a live video feed, with AI completing the job. But that may change soon. Drones may choose their own targets.

The balance between technological progress and the ill effects of AI is a policy debate that challenges governance around the world. Generative AI is really dangerous because its content can be misleading. An interesting conversation between the tech industry and senators in the Senate on May 16 (available on YouTube) is educational. As one senator pointed out, what and how we define technology is a key policy dilemma.

Georgia Senator John Ossoff asked Altman for his thoughts on this definitional conundrum. Altman helpfully suggests different regulatory thresholds based on “the amount of computing put into the model.” “A model that can persuade, manipulate, and influence people’s beliefs will be a threshold,” he said. He is already one threshold for models that can “create novel biological agents.” He suggests that each capability threshold should have a different level or regulation, and that models with lower capabilities should be kept for open use.

Regulation affects constitutional rights such as privacy, equality, liberty and life. That leads to a very important constitutional debate between the violation of the state and the assertion of individual privacy. After all, AI is only possible because data is collected. In India, while the state has unilateral rights to collect and use data, it also gives itself the power to regulate private entities. Private bodies and individual citizens have access to certain protections and rights. For now, the uses and dangers of AI will capture the imagination of policy makers. To create thoughtful and constitutional regulation, we need to educate ourselves about technology.

The author is a Senior Counsel for the Supreme Court