AI and truth
AI will not be infinitely helpful, and there will be many pitfalls to face.
Every child will have an AI tutor that is infinitely patient, infinitely compassionate, infinitely knowledgeable, infinitely helpful.” These are the words of the web pioneer Marc Adreessen, writing about how “AI Will Save the World.”
Such optimistic claims are not new. The rise of the world wide web came with predictions that it would be “a natural force drawing people into greater world harmony.” As we know, the web provided unprecedented access to information, but did not lead to “greater world harmony.” It became a medium for misinformation, echo chambers, bullying and polarization.
With increasing adoption of AI, there will be benefits in many areas like medicine, enhanced traffic safety and environmental monitoring. Even so, AI will not be “infinitely helpful” and there will be many pitfalls to face.
A post-epistemic world
Eric Horvitz, Microsoft’s chief scientific officer, recently sounded an alarm that I think is prophetic. Horvitz wrote that AI is moving us closer to a “post-epistemic” world “where fact cannot be distinguished from fiction.” In short, AI will have an impact on truth.
Horvitz identifies one area where AI will challenge truth: “deepfakes.” Deepfakes use AI to create synthetic videos that can impersonate people. In a famous demonstration, researchers at the University of Washington posted a deepfake video of President Obama, making him say whatever they wanted. Seeing is no longer believing and “truth” can now be manipulated and fabricated.
Other researchers have begun to recognize that AI chatbots are trained with a particular worldview and users are subject to something called “latent persuasion.” Regular usage of AI chatbots can be like having a “Jiminy Cricket” on your shoulder, autocompleting your thoughts. Over time, such nudging can shape your opinions without you realizing it.
Another recent development is “astroturfing” – using AI to generate a fake campaign that gives the illusion of a grassroots movement. AI chatbots can be harnessed to post massive amounts of tailored content on social media with the purposes of capturing attention and manipulating people’s opinions. Some predict that astroturfing will increasingly distort truth and reality, posing a direct threat to democratic societies.
What is the antidote?
In the mid-twentieth century, C.S. Lewis described the pitfall of developing a “blindness” to certain truths by reading “only modern books.” His advice is “to keep the clean sea breeze of the centuries blowing through our minds” by “reading old books.” In our twenty-first century era, “modern books” are no longer the issue, but rather “modern media.” None of us will be immune to the “blindness” that may be caused by misinformation, latent persuasion and astroturfing.
What is the antidote? Romans 12:2 urges us, “Do not conform to the pattern of this world, but be transformed by the renewing of your mind. Then you will be able to test and approve what God’s will is – his good, pleasing and perfect will.” In our context, “not conforming to the pattern of this world” might require developing counter-cultural habits of sabbath and limiting our connection to the constant stream of AI-driven media. Secondly, we need to “renew our minds” with the “clean sea breeze” of older books including, of course, Scripture.
We shape our tools, but our tools can also shape us – including shaping our perception of truth. With AI advances that promise to be “infinitely helpful,” let us remain faithful to the truth.
