The Need for AI Safety and Safeguarding Against Algorithmic Harms

The recent controversy at OpenAI, which resulted in CEO Sam Altman being fired and then rehired four days later, has brought attention to concerns about the development of artificial general intelligence (AGI) and the need to prioritize catastrophic risks. OpenAI's success with products like ChatGPT and Dall-E has raised questions about whether the company is focusing enough on AGI safety. AI is already widely used in daily life, but many algorithms exhibit biases that can cause harm, and efforts are being made to recognize and prevent these harms. While the development of large language models like GPT-3 and GPT-4 is a step towards AGI, it's important to consider the potential biases that may result from their widespread use in school, work, and daily life. The Biden administration's recent executive order and enforcement efforts by federal agencies are the first steps towards recognizing and safeguarding against algorithmic harms, particularly in the context of identifying individuals who are likely to be re-arrested. The deployment of AI may not be about rogue superintelligence, but rather about understanding who is vulnerable when algorithmic decision-making is ubiquitous.