Our guest today is Robert Moran. He’s a futurist, a management consultant, and a pollster. He began his career in political consulting and polling. He recently wrote a fascinating article at Campaigns & Elections about a thought experiment where an AI runs for office.
In our conversation, we dive deeper into his vision of an AI politician, what it would mean for our politics, and more near-term impacts of AI in campaigns.
00:00 - 00:08
In this context, Eric Wilson, the managing partner of Startup Caucus, discusses the weird future of campaign tech innovation on the right. He introduces himself and his role at Startup Caucus, which is the home of campaign tech innovation on the right. Although the future may seem strange, Wilson believes that it is something we will experience. The context provides a glimpse into the evolving landscape of political technology and sets the stage for further exploration of this topic.
00:08 - 03:20
In this podcast episode, Eric Wilson interviews Robert Moran, a futurist, management consultant, and pollster, about the potential of AI in politics. Moran explores the concept of an AI politician and the implications it could have on our political landscape. He discusses the rapid advancement of AI technology and how it may evolve to become artificial general intelligence, capable of independent thinking. Moran also highlights the increasing integration of AI in our daily lives and argues that people may become more comfortable interacting with AI in the future. However, he acknowledges the ethical considerations and cautions against electing a generic superintelligence. Instead, Moran suggests that basing an AI politician on a historical figure with predictable behavior and worldview could be a more viable option.
03:20 - 07:44
This context discusses the challenges and considerations involved in training an AI candidate for political campaigns. The author uses Abraham Lincoln as an example, highlighting the abundance of information available on him through speeches, letters, and biographies. The concept of bringing historical figures back to life through artificial general intelligence (AGI) is explored, acknowledging the difficulty of understanding their views on modern issues. The author suggests forming a team of historians to speculate on how the politician would conceptualize contemporary problems. The context concludes by mentioning the challenge of getting an AI candidate on the ballot and running a campaign.
07:44 - 11:25
In this context, the speaker discusses the challenges and possibilities of having an AI as a political candidate. Current laws require candidates to be citizens, but AI is not a citizen since it is not human. The only workaround would be to have a human proxy who promises to relinquish control to the AI upon election. However, this approach is fraught with problems, such as the proxy not fulfilling their promise. Additionally, there are concerns about foreign influence and the difference between swearing an oath to the constitution versus following an AI. If an AI candidate were to be elected, it would raise questions about how voters, constituents, and special interests could lobby the AI. The speaker suggests that constructing arguments that appeal to the AI's logic would be the best approach. The speaker also highlights that an AI candidate could engage with millions of voters simultaneously, potentially outcompeting human politicians in retail politics. Overall, the idea of an AI as a political candidate poses interesting theoretical and practical challenges to the current political landscape.
11:25 - 14:33
In this context, Robert Moran discusses the potential impact of artificial intelligence (AI) on politics. He predicts that AI chatbot avatars for real candidates could revolutionize the way we do politics. These avatars would allow voters to interact with candidates online. Moran also raises concerns about the addictive nature of chatbots and the percentage of constituents who may become addicted to them. He is currently working on a book called 'Lincoln 2.0' about the subject and hopes to have it published before AI candidates become a reality.
14:33 - 18:24
In this context, the speaker discusses the potential milestones and implications of AI in politics. They predict that by 2024 or 2026, AI candidates for office will emerge, although initially as large language models for publicity stunts. Presidential libraries are expected to develop their own presidential AIs using historical materials. In the long term, these AIs could be used as micro advisers for presidents. The speaker also mentions the possibility of achieving artificial general intelligence by the 2040s or 2050s, leading to rapid advancements and inventions. They emphasize the importance of exploring these ideas in advance and considering the ethical implications. The conversation also touches on the use of speculative fiction as a tool for envisioning future states and addressing ethical dilemmas. Lastly, the speaker raises the question of what laws and regulations should be in place for AI in politics.
18:24 - 22:01
The regulatory and legal framework for AI in politics is a complex and challenging issue. While some argue for minimal regulations to avoid slowing down progress, others believe that regulations are necessary to address potential risks and challenges. One area of concern is the use of deepfake technology, which can create realistic fake videos or audio recordings of individuals. This poses a risk to identity verification and could have significant implications, including in financial transactions and stock market events. The SEC and other regulatory bodies may need to develop strategies to mitigate these risks. Additionally, nation-states could exploit AI and deepfakes for information warfare. The speaker acknowledges the need for solutions but does not offer specific answers. They also mention the potential benefits of AI in politics, such as electoral form reforms like ranked-choice voting.
22:01 - 27:21
The context discusses the potential benefits of using AI in political campaigns and electoral reforms. It explores the concept of ranked choice voting, direct democracy, and liquid democracy as possible reforms that could enhance the democratic process. The conversation also delves into the idea of using AI to educate voters, inform them about legislative matters, and remind them to vote. Moreover, it highlights the near-term political use cases of AI, including creating AI-generated ads, avatars for candidate interactions, and training language models for effective fundraising appeals. The context concludes by referencing the movie 'Magic Town' and the use of digital agents to simulate citizens for conflict resolution purposes.
27:21 - 30:08
In the near future, political campaigns may utilize digital agents to test and optimize their strategies. These digital agents, which can number in the millions, would simulate different scenarios and run simulations to determine the most effective campaign formulations. This approach, similar to what sports teams and market researchers currently do, could have a significant impact on the political industry. However, the challenge lies in accounting for the unpredictable and irrational nature of human voters. While some are skeptical about the feasibility of modeling human behavior, the use of digital agents in political campaigns is an area of ongoing exploration.