AI Has Become a Technology of Faith

🌈 Abstract

The article discusses the grand claims and ambitions surrounding generative AI, particularly in the context of a new company called Thrive AI Health co-founded by Sam Altman and Arianna Huffington. The article explores the challenges and concerns around using generative AI for personal health and medical applications, and questions the ability of the technology to reliably and safely handle sensitive health data.

🙋 Q&A

[01] The Ambitions of Thrive AI Health

1. What are the key claims and ambitions of Thrive AI Health as described in the article?

• Thrive AI Health aims to bring OpenAI's generative AI technology into the healthcare domain, with the goal of improving health outcomes, reducing healthcare costs, and significantly reducing the effects of chronic disease worldwide.
• The centerpiece of their pitch is a "hyper-personalized AI health coach" that will generate personalized insights and recommendations based on a user's biometric and health data.
• Altman and Huffington compare their efforts to the New Deal, describing Thrive AI Health as "critical infrastructure" in a remade healthcare system.

2. What are the author's main concerns and criticisms about Thrive AI Health's ambitions?

• The author finds the ambitions "grandiose" and questions how realistic they are, given that the product is still conceptual with no clear details on what it will actually look like.
• The author is skeptical about the ability of generative AI to reliably handle sensitive health data and provide accurate medical advice, given the technology's propensity for hallucinations and inaccuracies.
• The author raises concerns about potential privacy and data misuse issues, such as insurers or data brokers obtaining users' personal health information.

[02] Generative AI and the "Faith-Based" Tech Industry

1. How does the author characterize the broader conversation around generative AI's capabilities?

• The author argues that much of the discussion around AI's greatest capabilities is based on a "vision of a theoretical future" and a "sales pitch" where current limitations are brushed aside.
• The author suggests that AI companies and researchers maintain a "rallying cry that the models 'just want to learn'" - the idea that with enough resources, the models will become increasingly capable.
• The author sees this as a "faith-based" approach to technology, where people are asked to judge AI not by its current limitations but by what is imagined for the future.

2. What are the author's concerns about this "faith-based" approach to AI?

• The author argues that this faith-based approach gives those who stand to profit from AI an "enormous amount of leverage" and opens the door for "delusion and for grifters looking to make a quick buck."
• The author suggests that this approach allows the industry to constantly "move the goal posts" and resist evaluation and criticism, as the promise of something glorious in the future continues to "string unwitting people along."