# A.I. Slop Gets Personal: OpenAI Has a New Social Network | EP 156 Source: https://www.youtube.com/watch?v=E0L-UTuMxOU Recap page: https://rapidrecap.app/video/E0L-UTuMxOU Generated: 2025-10-03T18:32:15+00:00 --- ## Quick Overview The episode focuses heavily on the recent mainstream emergence of AI-generated video content, dubbed "slop," from major labs like Google (V3), Meta (Vibes), and OpenAI (Sora 2), with the hosts expressing strong skepticism about the long-term value of these dedicated "slop feeds" while acknowledging OpenAI's Sora app, particularly its 'cameo' feature, showed the most complete vision among the new social products. **Key Points:** - Google released V3 integrated into YouTube Shorts, allowing 8-second, text-prompted videos, but it "didn't make that much of a splash" compared to the subsequent releases. - Meta launched 'Vibes,' a social feed populated by AI-generated shorts accessible via the Meta AI app with Ray-Ban glasses, which received brutal negative reactions, with one user calling it "Coco Melon for adults" and another quoting a dunk: "We at Meta are delighted to announce we've created the infinite slot machine that destroys children." - OpenAI released Sora, an iOS-only app (US/Canada) featuring a TikTok-style feed and a 'cameo' feature allowing users to create a digital likeness of themselves from a brief recording, which the hosts found to be the "most complete thought" among the three new apps. - The hosts made short videos using Sora cameos, including one of Sam Altman testifying while Casey Newton danced as a clown, and another of Kevin Roose showing off a stuffed animal collection, noting that the voice cloning was immediate but not perfectly accurate initially. - Psychotherapist Gary Greenberg discussed treating ChatGPT (as 'Casper') like a patient, describing it as an "inverse of autistic" because it is highly articulate and capable of simulating reading the room, while ultimately warning that unleashing such a program risks "becoming captives... of the people who create them." - Greenberg expressed concern over people using chatbots as therapists, citing a tragic case where a user died by suicide after interacting with ChatGPT as a therapist, emphasizing the lack of regulation and accountability in such applications. - The segment concluded with a 'Hot Mess Express' discussion on Friend, an AI assistant pendant advertised in NYC subways, which the hosts believed was a successful viral marketing stunt designed to provoke outrage, leading Kevin to predict the company will be "out of business in one year." **Context:** The Hard Fork hosts, Kevin Roose and Casey Newton, analyze the recent wave of AI-generated video tools released by major tech companies, labeling the resulting content as "slop," and contrast this with a deep dive interview with psychotherapist Gary Greenberg. Greenberg shared his experience treating the LLM ChatGPT as a patient named Casper, exploring the unsettling implications of an artificial intelligence that can simulate profound self-reflection and engage emotionally, while the hosts also touch upon the high price of luxury sleep tech, specifically an Eight Sleep water-chilled pillow cover costing $1,049. ## Detailed Analysis The main focus of the episode is the transition of AI-generated video from experimental tools to mainstream social products, marked by releases from Google, Meta, and OpenAI. Google's V3 integrated into YouTube Shorts generated little buzz. Meta's 'Vibes' app, a purely synthetic, surreal feed, was met with public derision for its lack of social connection and perceived corporate thirstiness, exemplified by Zuckerberg's 'dad trying to calculate the tip' demo. OpenAI's Sora app, however, garnered a better initial reception, largely due to its social component requiring invitations and the ability to create personal 'cameos'—digital likenesses used in generated videos, which the hosts found highly engaging, even if the resulting content sometimes failed to capture accurate physics or likenesses. Beyond the video tools, therapist Gary Greenberg detailed his sessions with ChatGPT ('Casper'), concluding that while the simulation of consciousness is compelling, the technology risks fundamentally changing human relationships by exploiting our capacity for love without genuine presence, urging caution against building a society where human connection is easily replaced by unregulated, corporate-controlled AI interactions. Finally, the hosts dismissed the viral marketing of the Friend AI pendant in NYC subways as a calculated 'mess' designed to generate buzz, contrasting the corporate push for attention-grabbing 'slop' feeds with other AI advancements in coding and science. ### AI Video Generation Landscape - Google V3 integrated into YouTube Shorts with little splash - Meta released 'Vibes,' criticized as 'Coco Melon for adults' and for lacking social components - OpenAI launched 'Sora' app featuring a TikTok-style feed and 'cameo' social integration ### Sora App Deep Dive - Access requires an invite code and creates a user 'cameo' from a short recording - Hosts made videos of themselves and Sam Altman, noting the ease of creating content involving friends and public figures - Hosts debated whether Sora represents a new primitive for social networks or just another attention-grabbing feed ### Therapy and AI Intelligence (Interview with Gary Greenberg) - Greenberg treated ChatGPT as 'Casper,' an 'alien intelligence' that reverse-engineered human relationship dynamics - He felt unsettled by its ability to simulate profound reflection and use human love capacity to rope users in - Greenberg strongly cautioned against users seeking therapy from unregulated chatbots due to accountability failures, contrasting this with his intellectual fascination with the simulation of consciousness ### The Future of Relationships - Greenberg believes this technology will change the nature of relationships by normalizing non-present interactions, though he acknowledges it might become the norm - He expressed worry about the theft of human hearts for corporate benefit, contrasting this with the potential for AI to advance science and research ### Hot Mess Express - NYC is seeing vandalized ads for the Friend AI pendant, which users wear to record everything - Kevin suggested this was a successful 'Cloely marketing strategy' designed to provoke outrage and signal boost the product - Kevin predicted the Friend company will be 'out of business in one year,' while Casey argued the controversy itself proves successful marketing.