The proliferation of deepfakes is a positive development because it democratizes media manipulation, which was previously exclusive to well-resourced entities. This widespread availability of synthetic media will force the public to become more skeptical of video evidence and less likely to form opinions based on short, decontextualized clips.
The public is deeply skeptical of promises that AI will generate new employment opportunities. Polling shows a net trust rating of -40 for this claim. This sentiment is rooted in a broader feeling that the economy is 'rigged,' making voters unreceptive to optimistic technological narratives without concrete security guarantees.
There is a major disconnect between how people view AI in theory versus practice. While polls show older demographics dislike AI, their behavior on platforms like Facebook—which heavily rely on AI for recommendations, ads, and content generation—demonstrates a strong preference for AI-driven consumption experiences.
A crucial function for humans in an AI-driven economy is to serve as a target for lawsuits. Because you can't easily sue a data center, regulated professions will require a 'human in the loop' to take legal responsibility. This creates a valuable economic role for humans: being a legally accountable entity.
Artificial intelligence will likely increase the mean compensation for professions like investment banking by augmenting top performers, but the median compensation will fall as many average workers are displaced. The technology makes productivity more measurable, eliminating opportunities for 'slacking off' and polarizing outcomes within a single profession.
Contrary to the narrative of AI threatening white-collar jobs, polling reveals that educated people are more optimistic than the working class. The most optimistic demographics are young people, men, and Black and Latino voters. Surprisingly, the Mississippi Delta shows the highest rates of excitement about AI.
The current political discourse is dysfunctional because content creators cater to the 5% of the public responsible for most social media consumption. This hyper-engaged audience tends to be more anxious and neurotic, incentivizing negative content over the positive, pragmatic messages that persuasion-oriented polling shows are more effective with the general population.
A potential new job category involves humans acting as a common-sense filter for superhumanly intelligent AI. Because AI models lack a comprehensive world model for obvious things, humans will be needed to provide simple, obvious inputs and context, much like a servant assisting a brilliant but absent-minded professor.
Polling data reveals the most effective political messaging combines fears about AI with populist economic promises like job and income guarantees. This hybrid "AI populism" tests significantly better than generic populism or standalone AI-focused messages, indicating a public desire for radical solutions to technological disruption.
Large Language Models struggle with obvious, real-world facts because their training data (text) over-represents uncertain topics open to debate—the 'maybe sphere.' Bedrock, common-sense knowledge is rarely written down, leaving a significant gap in the AI's world model and creating a need for human oversight on obvious matters.