Lesson Report:
Title
From “Daisy� to the Feed: How Micro‑Targeting Rewires Political Persuasion
In this session, students traced the evolution from broad-reach political advertising to data-driven micro-targeting. Using the 1964 “Daisy� ad and a contemporary targeted ad scenario, they analyzed emotional appeals, media environments, and assumptions about a “shared public conversation.� The objective was to understand how algorithmic targeting leverages personal data and anxieties, and to surface the democratic risks such systems create.
Attendance
– Present (reported by instructor): 23 participants
– Absent explicitly mentioned: 0
– Notes: One student (Sajid) joined the course for the first time and late due to access issues.
Topics Covered (chronological)
1) Framing: AI beyond LLMs; algorithms and advertising
– Instructor clarified a gap from a prior session: AI is not just chat-based LLMs; recommender/optimization algorithms that power feeds and ads are a foundational AI application that enabled today’s LLM ecosystem.
– Key concept defined: Micro-targeting in advertising
– Old model: Spend broadly (billboards/TV) → large reach, low relevance.
– New (algorithmic) model: Collect user-level data → infer preferences, vulnerabilities → deliver hyper-relevant ads likely to trigger emotional and behavioral responses (buy, click, vote).
– Goal for the day: Show how micro-targeting works on products, ideas, and candidates; why it’s powerful and potentially dangerous.
2) Historical case study setup: 1964 U.S. context
– Brief Cold War context: Post–WWII U.S.–U.S.S.R. standoff, nuclear arms race, pervasive public fear of nuclear war.
– Electoral context: Lyndon B. Johnson (Democrat) vs. Barry Goldwater (Republican). Note: Instructor corrected a mid-class memory slip and confirmed LBJ was the Democratic candidate; Goldwater the Republican.
– Preview of analysis: Students were asked to attend to visuals, audio, text, and the ad’s “vibeâ€�/emotion.
3) Activity 1: View and analyze the “Daisy� ad (1 minute)
– Logistics: YouTube link shared in Zoom; class asked to confirm audio/video. Plan was to watch twice (first for atmosphere; second for analysis).
– Breakout instructions (5 minutes, groups ~4 students):
1) Intended audience: Who was meant to see this?
2) Core emotional message: What should viewers feel?
3) 1964 reception and media environment: How might a citizen have perceived it? What does it reveal about what was acceptable/normal in media at the time?
4) Assumptions about a shared public conversation: Did it exist? Where/how did it occur?
4) Debrief: Findings from the “Daisy� ad
– Intended audience
– Broad U.S. electorate, especially undecided voters and family-oriented viewers; anyone anxious about nuclear war.
– Targeting was wide because nuclear annihilation is a near-universal fear; leverages maximal emotional salience.
– Core emotional message and tactics
– Primary emotion: Fear—of nuclear destruction and of an immediate, catastrophic consequence from making the “wrongâ€� political choice.
– Secondary themes: Urgency (“stakes are too highâ€�), morality/religion (good vs. atheist/immoral adversary subtext typical of Cold War rhetoric), innocence/endangered children (heightened empathy).
– Implied contrast: Johnson = safety/responsibility; Goldwater = danger/escalation.
– 1964 reception and media environment
– Students noted the ad’s unusual shock value for the era: television was relatively new in households; graphic content (implied child death via nuclear blast) was far less common than today.
– Instructor fact: The “Daisyâ€� ad aired once and was pulled after intense negative reaction. It became infamous for overplaying manipulative fear appeals.
– Media landscape characteristics: Few national TV channels; centralized, top‑down broadcast with minimal viewer control; high common exposure to the same content.
– Assumptions about a public conversation (then)
– Conversation flows top‑down (broadcasters/political campaigns → public). No “comment sectionâ€�; feedback occurs offline in fragmented, interpersonal settings (barbershops, classrooms, family gatherings).
– Relies on a shared cultural anxiety (nuclear war) and a relatively unified media diet to seed mass, cross‑group discussion.
– Assumes citizens attend to “big‑Pâ€� politics (national stakes, national leadership) and will integrate shocking messages into everyday talk.
5) Transition: From mass broadcast to micro‑targeting
– Instructor contrasted 1964’s wide-cast “one message to allâ€� with today’s individualized feeds that exploit personal data.
6) Activity 2: Build a micro‑targeting profile (“Anna�)
– Materials: A data-broker-style profile for “Annaâ€� (Google Doc link shared).
– Task (5–6 minutes in groups): Infer
1) Who Anna is (demographics/situation).
2) What she worries about (with evidence from the profile).
3) What ideas/frames engage her (and why).
– Hinted data categories (as shared in class discussion): Age ~20s; Bachelor’s in Business Administration (2018 grad); salary reported at ~$55,000/year; follows/uses budgeting apps; active on LinkedIn/Indeed; engages with content about layoffs, cost of living, housing affordability; first‑time homebuyer and alumni groups; burnout/self‑care content.
7) Debrief: What the class inferred about “Anna�
– Identity/situation
– Recent(ish) college graduate, early‑career professional with modest income for U.S. norms; likely urban or suburban; pursuing career advancement (heavy LinkedIn/Indeed use).
– Salient anxieties
– Economic insecurity: inflation/groceries, layoffs, precarious employment, housing affordability; probable job search or job change.
– Emotional state: elevated anxiety/stress; engages with burnout/self‑care content, signaling coping needs.
– Ideas/frames that engage her
– Financial literacy/budgeting, economic trend explainers, career mobility tips, first‑time homebuyer guidance, peer communities (alumni/homebuyer groups).
– Messaging that acknowledges material constraints and offers concrete, low‑friction steps resonates (e.g., “beat rising costs,â€� “protect your future,â€� “land a stable jobâ€�).
– Instructor synthesis: With relatively few data points, advertisers can infer actionable psychological and situational traits (e.g., money worry, job hunt status) that are exploitable for persuasion.
8) Activity 3: Contemporary political micro‑ad targeted to “Anna�
– Ad shown (as a YouTube mid‑roll): “They say Freshta is bringing jobs back, but you know the truth. Your future is on the line. Vote for Anush.â€�
– Delivery strategy: Shown to Anna 6 times over 90 days (repetition to reinforce salience/availability; aligns with her feed behavior and search history).
– Rapid analysis prompt: How has the “public conversationâ€� assumption changed since 1964?
– Student observations and instructor synthesis:
– From shared mass conversation to fragmented, personalized “micro‑publics.â€� Each user sees a curated slice; far less common knowledge or shared reference points.
– Ad content exploits personal anxieties (economy/jobs) surfaced by data trails, not society‑wide fears.
– Frequency and timing are optimized to the individual—sustained nudges rather than a one‑time “eventâ€� ad.
– Feedback loops exist (likes/comments/shares), but occur within segmented networks; controversy seldom spills into a unified, national conversation.
– Takeaway contrast:
– 1964: One dramatic ad aimed at everyone to seed broad, offline debate.
– Today: Many tailored ads aimed at specific personas to quietly shift individual attitudes/behavior without triggering broad scrutiny.
9) Wrap‑up and forward look
– Today’s focus: What micro‑targeting is, how it works, and why it can be democratically risky.
– Next session: Why and how micro‑targeting became legally/structurally possible—examining the legal architecture and the weakening of democratic “guardrails.â€�
Actionable Items
Urgent (before next class)
– Post assigned reading on eCourse: Chapter on legal constructions that weakened democratic guardrails (as promised).
– Centralize links from class:
– 1964 “Daisyâ€� ad (YouTube).
– “Annaâ€� profile Google Doc used in Activity 2.
– WhatsApp group access: Ensure a working invite link is distributed and confirm all students are in the group (instructor had trouble generating link; Sahar shared one—verify it works).
– Correct the record in posted notes: LBJ was the Democratic candidate in 1964; Goldwater the Republican.
High Priority (this week)
– New student onboarding:
– Provide Sajid with lesson notes to date and confirm access to eCourse/links.
– Offer office hours walkthrough to bring him up to speed.
– 1:1 follow-ups:
– Confirm/complete the planned Google Meet with Safi and note any required follow-on actions.
– Tooling clarity:
– Decide and communicate a policy on the Zoom “AI Companionâ€� (consent, recording/transcript use, privacy).
Nice to Have
– Supplementary materials:
– Post a short background note on the “Daisyâ€� ad’s broadcast history and reception, and a brief primer on Cold War nuclear drills to enrich context.
– Pedagogy/logistics:
– Calibrate breakout timing and group sizes given 23 attendees; consider pre-assigned groups to speed transitions.
– Prepare a brief explainer on ad delivery mechanics (frequency capping, audience construction, lookalike modeling) to support next session’s legal/policy discussion.
Homework Instructions:
ASSIGNMENT #1: Reading — Legal foundations of micro-targeting and the weakening of democratic guardrails
You will read the chapter the professor is posting to help you understand how legal and policy choices made today’s micro-targeting possible, building directly on our comparisons between the 1964 “Daisy� ad (mass broadcast) and the personalized, anxiety-focused ad repeatedly shown to “Anna� (algorithmic targeting). This prepares you to discuss how assumptions about a shared public conversation have shifted and why.
Instructions:
1) Access the assigned chapter:
– Go to eCourse and open the reading the professor is posting for this week on the legal constructions that weakened democratic guardrails.
– If you do not see the reading, message the professor promptly.
2) Read actively and annotate:
– As you read, highlight definitions, legal mechanisms, and historical milestones the author identifies as enabling targeted advertising and data-driven campaigns.
– Note any examples the author uses that resemble the “Annaâ€� profile case we discussed (data brokers, platform signals, repeated exposure to tailored ads).
3) Use these guiding questions to focus your notes (for your own use in discussion):
– Which “democratic guardrailsâ€� does the author say have weakened, and how are they defined?
– What legal or regulatory choices (or gaps) allowed platforms and campaigns to collect and use individualized data at scale?
– How do these choices change assumptions about a shared public conversation compared to 1964 broadcast TV (few channels, common content) versus 2025 feeds (fragmented, personalized streams)?
– How does micro-targeting leverage emotions (e.g., fear and urgency) differently today, and why does the law matter for that difference?
– What remedies or reforms, if any, does the author propose?
4) Connect the reading to our class examples:
– Revisit the contrasts we drew:
• 1964 LBJ “Daisy� ad: broad, one-to-many message relying on a common national media environment.
• The “Anna� scenario: one-to-one or one-to-few targeting using platform data (job anxiety, cost-of-living stress) and repeated frequency capping (shown 6 times in 90 days).
– Be ready to explain how the chapter’s legal analysis helps make sense of that shift.
5) Prepare for Tuesday’s discussion:
– Bring your annotated notes (digital or paper) so you can cite specific passages and concepts from the chapter.
– No upload is required; you will use your notes to contribute to the in-class conversation at the start of the next session.
6) Time management tip:
– Allocate enough time to read carefully and annotate (plan for at least one focused sitting), so you can participate meaningfully in discussion.