MIT Professor Warns Congress of AI’s Critical Turning Point

During a recent hearing before the House Subcommittee on Cybersecurity, Information Technology, and Government Innovation, Aleksander Mądry, the Cadence Design Systems Professor of Computing at MIT and director of the MIT Center for Deployable Machine Learning, called on the government to play an active role in regulating the use of artificial intelligence (AI). Mądry argued that it is not sufficient to leave the development and deployment of AI solely in the hands of large tech companies.

Mądry highlighted the importance of government oversight in ensuring that AI is developed and used in ways that are consistent with society’s goals. He called on the government to ask important questions about the purpose and explainability of the algorithms used in AI, as a precursor to regulation. Without such oversight, Mądry warned that he is “”extremely worried”” about the future of AI.

Mądry also discussed the risks associated with AI, including the potential for AI to interact in ways that go against human intuition. He urged caution in assuming that large language models like ChatGPT produce credible results. Mądry also discussed the problems that can arise in the AI “”supply chain”” – the way AI systems are built on top of each other – and the difficulty in holding someone responsible when a problem arises.

Mądry concluded his testimony by calling on Congress to act quickly to address the issues surrounding AI regulation. He urged policymakers to engage in a difficult conversation about the role of AI and how to ensure that it benefits everyone.