David Rand

Overview

David Rand is a Cornell researcher cited by Marketplace Tech for experimental work on political persuasion through conversational AI.

Current Profile

In the episode, Rand describes a purpose-built chatbot tested before the 2024 U.S. presidential election. An average conversation lasted about six minutes and shifted some participants’ stated preferences between Kamala Harris and Donald Trump in both directions. He characterizes the measured movement as roughly three to four times the effect associated with watching a standard political advertisement.

Rand’s role in the wiki is therefore narrow and evidence-bound: he supplies a research claim that interactive, information-rich dialogue may produce stronger short-term political persuasion than a conventional ad. The episode does not provide enough methodological detail to establish durability, population-level electoral impact, or equivalence to ordinary use of ChatGPT, Claude, or Gemini.

Key Characteristics

• Rand studies how conversational systems can affect political attitudes.
• His reported 2024 experiment used a purpose-built chatbot rather than an unspecified commercial assistant.
• The reported movement occurred toward both major presidential candidates rather than in only one partisan direction.
• He interprets interactive access to substantial policy information as one possible reason for the measured effect.
• His result is presented as a comparative experimental effect, not as proof of real-world election outcome change.

Evidence

• Experimental setting - Using generative AI for election-related research identifies Rand as a Cornell professor discussing a chatbot his team built before the 2024 U.S. presidential election.
• Direction and duration - Using generative AI for election-related research reports an average conversation of about six minutes and preference movement in both candidate directions.
• Comparative effect - Using generative AI for election-related research attributes to Rand an effect estimate roughly three to four times that of a standard political advertisement.

Qualifications

The source note summarizes a short interview rather than the underlying study. It does not establish the sample, randomization, exact intervention, statistical uncertainty, persistence of attitude change, voting behavior, or generalizability to commercial chatbots. Institutional affiliation and numerical comparisons are retained as episode-reported claims.

What Changed

• Created a bounded profile for Rand’s reported conversational-AI persuasion research.

Relationships

• Cornell University - institutional affiliation reported by the episode.
• Conversational AI Political Persuasion - research topic for the chatbot experiment he describes.
• Kamala Harris - one candidate toward whom some participants reportedly moved.
• Donald Trump - the other candidate toward whom some participants reportedly moved.
• Marketplace Tech - program carrying the interview excerpt.