About Personhood and the Perception of Truth in the Age of AI
| Authors | Schuurman, Derek |
|---|---|
| Date | 2024 |
| Place | Gordon College |
| URL | https://www.youtube.com/watch?v=BGO6VgDIK4k |
| Language | en |
Type: Video Recording
Tags: Artificial intelligence, Chatbots, Christian perspective, computers, creation, direction, Ethics, fall, functionalism, Humanity, kingdom of God, large language models, Macintyre, personhood, Postman, redemption, reductionism, Reformational worldview, statistics, structure, Technology, truth, Turing test, Weizenbaum, Wolters, Worldview
Abstract
The talk began with a Dr. Seuss-like poem about Gordon College to illustrate what today's chatbots can produce with data from the web. Statistical plausibility is the goal, not truth; in fact, AI is sometimes used for propaganda or disinformation, to persuade people of something that's not true. Large language models are intrinsically grounded in a worldview. This comes out in chatbots adopting personas, when asked, such as being C. S. Lewis. AI tutors are now being deployed for students, AI pastors can be called up, as can AI friends, AI Jesus, etc. AI is starting to replace human relations and contacts and shape who we are. AI is not new; the ideas go back to Turing in the 1950s, who posited the Turing Test for judging when AI is successful - when we can't tell the difference between a machine and a human being in its responses. The central question that arises is, what does it mean to be human? What credentials do humans have? What is personhood? Ethical considerations and worldview commitments are involved in the use of AI. A Christian use should be grounded in the Biblical worldview of Creation, Fall, Redemption. AI is a latent part of God's creation that can be used in fallen ways or in ways that are redemptive. AI, like all technology, can be used in ways that help build God's kingdom. Christians should pay attention to design norms as they develop and use AI. This will include not treating chatbots as if they are persons and developing thoughtful regulations for AI so it is used appropriately. [CJ]