Listen and learn.

A philosophical heuristic distinguishing four purpose-relative levels at which synthetic media may represent a person. It explains how a deepfake can humiliate without fooling viewers; it is not a detector or a universal legal rule. Source citation: Carl Öhman. “The identification game: deepfakes and the epistemic limits of identity.” Synthese 200, 319 (2022). https://doi.org/10.1007/s11229-022-03798-5.

A thematic analysis of 1,231 tweets from January–August 2022. It identifies three forms of deepfake discourse and sample-level examples of real footage being called fake; results are not a representative measure of public trust, and the authors call for broader research. Source citation: John Twomey, Didier Ching, Conor Linehan, Matthew Peter Aylett, Michael Quayle & Gillian Murphy. “Do Deepfake Videos Undermine Our Epistemic Trust? A Thematic Analysis of Tweets During the Russo-Ukrainian War.” PLOS ONE 18(10), e0291668 (2023). https://doi.org/10.1371/journal.pone.0291668.

A comparative analysis of 90 Miles and All Eyes on Rafah, exploring privacy, indexical evidence, iconic representation, affect, and solidarity. The authors do not predetermine either case as witnessing; the controversies themselves are the subject. Source citation: Phoebe Matich, Michael Richardson & Jean Burgess. “Authenticity controversies: Witness media and generative AI imagery.” Media, Culture & Society, OnlineFirst (2026), pp. 1–19. https://doi.org/10.1177/01634437261442039.

An interview-based analysis of In Event of the Moon Disaster and Gerry Anderson: A Life Uncharted, focused on synthetic voice and face, disclosure, consent, and documentary ethics. This is a two-case production study, not a general audience survey. Source citation: Dominic Lees. “Deepfakes in documentary film production: images of deception in the representation of the real.” Studies in Documentary Film (2024). https://doi.org/10.1080/17503280.2023.2284680.

A practice-informed study of generative AI as asset, tool, and collaborator in computational non-fiction, including four projects and the authors’ discussion of MEL 23:55. It examines creative process rather than audience effects or model accuracy. Source citation: Max Schleser & Susan Kerrigan. “Examining the emergence of the ‘AI eye’ and its effect on the ‘creative treatment of actuality’ in computational non-fiction.” Media Practice and Education 26(4), 444–460 (2025). https://doi.org/10.1080/25741136.2024.2380757.

A source-grounded summary of a WITNESS report using seven adversarial scenarios to examine how provenance infrastructure could expose or track creators, sources, and viewers. These are risk-analysis scenarios, not evidence that every abuse has occurred or an official C2PA position. Source citation: Jacobo Castellanos Rivadeneira. “C2PA Content Credentials and the Surveillance Risk: Adversarial Scenarios and Governance Gaps in the Content Provenance Ecosystem.” WITNESS (2026). https://library.witness.org/product/c2pa-privacy/.

A reader’s guide to C2PA Technical Specification 2.4: manifests, claims, signatures, asset bindings, and validation. This is a technical standards overview, not a scholarly study; cryptographic validity is not proof that a depicted event is true. Source citation: Coalition for Content Provenance and Authenticity (C2PA). Content Credentials Technical Specification, version 2.4, revision date 2026-04-01. https://spec.c2pa.org/specifications/specifications/2.4/specs/C2PA_Specification.html.

A source-grounded summary of Fallis’s philosophical argument that deepfakes threaten knowledge both by deceiving viewers and by weakening video’s evidentiary value. This is a conceptual analysis, not a survey or a claim that all video is suspect. Source citation: Don Fallis. “The Epistemic Threat of Deepfakes.” Philosophy & Technology 34(4), 623–643 (2021). https://doi.org/10.1007/s13347-020-00419-2.

A research summary of Nick Hector’s 2026 article, based on interviews with twelve experienced documentary editors. It examines how transcription, search, and related AI tools may support editing—and why the interviewees warn that faster workflows could crowd out full viewing, craft, time, ethical judgment, and the apprenticeship pipeline. Source: https://doi.org/10.1080/25741136.2026.2705320.

A research summary of Sam Gregory’s essay on deepfakes, civic witnessing, and the systems built to establish media authenticity. The episode explains shallowfakes, the liar’s dividend, frontline and distant witnesses, provenance as a verification subsidy, and the risks of surveillance, exclusion, and unequal access when authenticity infrastructure becomes a condition of being believed.

A research summary of Feng, Ritchie, Blumenthal, Parsons, and Zhang’s study of how provenance information changes viewers’ trust and perceived accuracy of visual media. The episode explains the 595-person experiment, why provenance can expose deceptive composites, how incomplete records can cause overcorrection, and what usable provenance means for documentary and AI-assisted media.

A research summary of Meghanne Barker’s argument that documentary authenticity is not limited to untouched camera footage. The episode explains animated authenticity, the limits of mechanical indexicality, the role of staging and reconstruction, and the ethical questions that arise when creative form helps a film communicate a higher truth. It closes by applying Barker’s framework to AI-assisted documentary: disclose intervention, protect the people represented, and make the relationship to reality legible and accountable.

A practical two-host introduction to IDEO’s Human-Centered Systems Thinking course: mapping complexity, listening to stakeholders, finding leverage points, and testing responsible change.

A faithful two-host summary of Cal Newport's YouTube video, How to Reinvent Your Life in 2026. The episode explains Newport's Deep Life Map: capability, values, and vision; how to identify the territory that needs attention; and how to make one concrete experiment that creates evidence toward a more purposeful life. Rendered with Qwen3-TTS using the consented cloned female and male reference voices.

Bill Gates just published a striking essay on Gates Notes about AI's speed, risks, and what a real plan to manage the transition could look like. We break down his three big risks (jobs, misuse, and kids' development), the genuine upsides in health, farming, and education, and his three-part plan: a new global framework, 'Human Reserved' jobs, and taxing AI and robots.

A beginner-friendly two-host introduction to Bowen family systems theory, its core concepts, evidence limits, and a practical way to notice patterns without diagnosing family members.

A practical two-host episode connecting Bowen family systems theory to communication, partnership, parenting, boundaries, leadership, and family life without turning the theory into a diagnosis.

A teaching-first two-host guide to Zuckerberg’s personal-AI vision: what a personal agent could do, where user control can slip away, and the privacy, permission, and user-power questions worth asking. Rendered locally with Qwen3-TTS cloned voices.

Two hosts unpack Pope Leo XIV's Magnifica Humanitas: human dignity, the common good, truth, education, work, family, freedom, war, and responsible artificial intelligence. A beginner-friendly guide grounded in the official Vatican text and rendered locally with Qwen3-TTS cloned voices.

A practical two-host guide to how applicant-tracking systems parse, search, and filter applications, plus advanced AI workflows for requirement matrices, verified evidence banks, ethical tailoring, strategic-stretch decisions, interview preparation, privacy, human referrals, and job-scam verification.

A beginner-first two-host guide to OpenAI’s Agents SDK: agents, tools, handoffs, guardrails, tracing, and practical ways it could support MissionCAP, learning programs, publishing, and research without over-automating your life. Rendered locally with Qwen3-TTS cloned voices.

A practical, beginner-first two-host conversation on using a personal Hostinger site and Hermes Agent to publish teaching, consulting, filmmaking, and community work—without turning life into a content machine. Covers public pages, podcast publishing, useful uploads, accessibility, privacy, maintenance, and a simple MissionCAP next step. Rendered locally with Qwen3-TTS cloned voices.

A beginner-first two-host lesson on Superlinked and its open-source SIE platform: embeddings, semantic search, reranking, extraction, RAG, self-hosting, privacy, small-model inference, and where the technology could fit in a MissionCAP or Hermes workflow. Rendered locally with Qwen3-TTS cloned voices.

A NotebookLM-style deep dive into NVIDIA's free serverless NIM API endpoints, what they mean for builders, and how to evaluate them responsibly. Rendered with Qwen3-TTS using the cloned female and male reference voices.

Part 2 of a two-part deep dive into Google's white paper on harness engineering, orchestration, production agent workflows, safeguards, economics, and adoption. Rendered with Qwen3-TTS using the cloned female and male reference voices.

Part 1 of a two-part deep dive into Google's white paper on the shift from vibe coding to agentic engineering, context engineering, verification, and the AI software factory. Rendered with Qwen3-TTS using the cloned female and male reference voices.

Two hosts explore the practical computer-science concepts and decision-making lessons from Algorithms to Live By by Brian Christian and Tom Griffiths. Rendered with Qwen3-TTS using the cloned female and male reference voices.

Two hosts examine motivated reasoning, calibration, and the practical lessons from The Scout Mindset by Julia Galef. Rendered with Qwen3-TTS using the cloned female and male reference voices.

Two hosts explore decisions under uncertainty and the practical lessons from Thinking in Bets by Annie Duke. Rendered with Qwen3-TTS using the cloned female and male reference voices.

Two hosts unpack the major game-theory concepts and practical lessons from The Art of Strategy by Avinash Dixit and Barry Nalebuff. Rendered with Qwen3-TTS using the cloned female and male reference voices.

A comprehensive two-host deep dive on the four major privacy frameworks governing medical information in Ontario, Canada, with practical compliance guidance for health custodians. Rendered with Qwen3-TTS using the cloned female and male reference voices.

A beginner-first introduction to OpenCV: what computer vision is, how images and video become useful answers, and how to build a simple red-ball follower. Once the fundamentals are clear, the two hosts explain the practical highlights and migration considerations in OpenCV 5. Rendered locally with Qwen3-TTS cloned voices.

A controlled eight-sample comparison using identical text and the same cloned female and male voices. Order: 1.7B full-reference female/male; 1.7B trimmed-reference female/male; 0.6B full-reference female/male; 0.6B trimmed-reference female/male. Average render times were 26.4, 20.5, 16.8, and 17.9 seconds respectively. All eight clips achieved effectively exact Whisper transcription; one clip had a single article substitution.

A beginner-friendly two-host lesson following Sunrise Breads from website files through previews, production deployment, domains, logs, rollback, accessibility, and responsible maintenance, rendered with Qwen3-TTS using the cloned female and male reference voices.

An evidence-grounded two-host episode about stable self-worth, self-compassion, balanced thought review, boundaries, repair, and reciprocal connection, rendered with Qwen3-TTS using the cloned female and male reference voices.

An evidence-grounded two-host episode about building domain-specific confidence through action, mastery experiences, useful feedback, fair interpretation, and calibration, rendered with Qwen3-TTS using the cloned female and male reference voices.