The author further discusses the concept of "botshit", a term used to describe untrue material created by generative AI that could be used uncritically and thoughtlessly, making it harder for people to discern truth from falsehood. The article warns of the dangers of AI-produced "hallucinations" being used to answer important but difficult-to-verify questions, potentially leading to voters making decisions based on illusory information. However, the author suggests measures to limit these dangers, such as watermarking AI-generated content, training AIs on authoritative information sources, and encouraging voters to exercise critical judgment.
Key takeaways:
- Generative AI could profoundly reshape politics, with the potential for false information being created at little or no cost and highly personalised advertising being produced to manipulate voters.
- Concerns about the impact of generative AI on elections have become urgent, with the technology likely to influence many major elections around the world in 2024.
- Generative AI technologies such as ChatGPT can produce what are called “hallucinations”, or artificially generated information that bears little relationship to reality, potentially leading to an environment where people make important voting decisions based on false information.
- There are measures that could be taken to limit the dangers of AI technologies, including watermarking AI-generated content, training AIs on authoritative information sources, and encouraging voters to exercise critical judgment by reality-checking important pieces of information.