Get Your Weekly Digest

Receive the best podcast summaries and insights delivered to your inbox every week.

We respect your privacy. Unsubscribe anytime.

My First Million

The App That'll Be Bigger Than TikTok

Published October 3, 2025
Visit Podcast Website

About This Episode

The hosts discuss OpenAI's new Sora app for AI-generated video, exploring its onboarding flow, social mechanics, and why it may be more powerful than TikTok as a multiplayer AI experience. They broaden the conversation to AI as a super app (including ChatGPT's Pulse), concerns about OpenAI's growing power, and how AI will reshape content creation, education, therapy, and addiction support. The episode also covers the rise of micro sports betting and prediction markets, new businesses tackling gambling addiction with AI, and the extreme personal data logging practices of Shopify CEO Tobi Lütke alongside their own approaches to life-logging and memory capture with tools like Meta smart glasses.

Topics Covered

Artificial intelligence User-generated content Gambling addiction and bookmaking Prediction markets Therapy Alcohol addiction and recovery Entrepreneurship Personal Development Mental health Online retail

Disclaimer: We provide independent summaries of podcasts and are not affiliated with or endorsed in any way by any podcast or creator. All podcast names and content are the property of their respective owners. The views and opinions expressed within the podcasts belong solely to the original hosts and guests and do not reflect the views or positions of Summapod.

Quick Takeaways

  • Sora is an AI video app with a TikTok-style feed that can clone a user's face and voice from minimal input and generate highly realistic videos, including social interactions through tagging.
  • The hosts see Sora as one of the first truly "multiplayer" AI products and predict it could become the fastest app ever to reach 100 million downloads once invites are opened.
  • They argue that ChatGPT's new Pulse feature and similar tools are early examples of AI "super apps" that could centralize news, learning, and productivity in a highly personalized feed.
  • Rapid AI progress leaves people feeling both newly powerful and potentially irrelevant, as AI can now produce content and even full podcasts that rival human-made media in watch time.
  • Micro sports betting is exploding, with many young men placing constant small wagers and facing aggressive incentives from platforms, contributing to rising gambling addiction.
  • New startups like Birches Health and Sunflower Sober use telehealth and AI to support people with gambling, alcohol, and other addictions, offering scalable alternatives or complements to traditional AA and rehab.
  • The hosts highlight how AI chat tools already function as informal therapists or executive coaches, helping them process emotions, conflicts, and communication dilemmas.
  • Shopify CEO Tobi Lütke has logged every keystroke and frequent screenshots of his computer for around 15 years, creating a unique personal dataset that could power an extremely accurate AI version of himself.
  • The hosts distinguish between tracking for optimization and archiving for memory, with one preferring targeted tracking for behavior change and the other focused on rich, long-form home video archives and potential life-logging.
  • Meta smart glasses are praised as a low-friction way to capture authentic family moments without pulling out a phone, and are seen as a powerful tool for future personal memory archives.

Podcast Notes

Introduction and First Impressions of Sora

Initial reaction to downloading Sora

Host describes Sora as an app that "blew my mind" and feels different from previous AI releases[0:00]
• He contrasts his reaction with typical AI hype where people say each new model has "changed the game" but this felt genuinely weird and unsettling in a new way

Basic concept of Sora

Sora is introduced as an OpenAI app with a TikTok-like feed where all videos are AI-generated[0:39]
• The host notes that you don't have to wonder if a video is fake because "they're all fake videos" by design
Initial internet backlash labeling Sora content as "AI slop"[0:53]
• He cites a tweet by Chris Bakke joking that OpenAI went from saying "we're going to cure brain cancer" to "we created brain cancer" with an AI video feed
• The host argues that dismissing Sora as "AI slop" may be a mistake and could put critics on the wrong side of history

How Sora Works: Onboarding, Face and Voice Capture

Unusual onboarding flow

Camera and microphone-based onboarding[1:39]
• When the app opens, the camera is immediately on and the user is instructed to say three numbers aloud, e.g., "12, 74, 88"
• Sora uses those three spoken numbers to capture and clone the user's voice, which the host describes as effectively "stole your voice"
• The app then asks the user to look to the side briefly, capturing facial data from a couple of simple actions to recreate the user's face
Implications of minimal data for full-body inference[2:46]
• Host remarks that Sora was able to infer his entire body, including apparent height, from just his face data, which he finds "disappointing in its own way"

Examples of Sora-generated videos using the host's likeness

Ralph Lauren ad and ponytail video[2:16]
• He created a video of himself in a "Ralph Lauren ad" as a playful test
• Another prompt was to "make it look like I have a ponytail," resulting in a realistic video of himself describing having grown his hair out and showing how he ties it back
Ding-dong ditch scenario[2:46]
• He generated a video of himself "ding-dong ditching" someone using simulated Ring doorbell footage, then doing a cartwheel to escape
• The realism of the motion and environment led the co-host to comment that it looks real

Using other people's likenesses in Sora

Sam Altman as a public template[3:21]
• The host explains that users can generate videos featuring Sam Altman if he is tagged, e.g., "me talking to Sam Altman"
• He describes a viral Sora video where Sam Altman addresses users as "my piggies enjoying their slop," mocking the idea of AI slop
Assessment of product quality relative to TikTok[3:49]
• The host claims Sora is "100%" better than TikTok and emphasizes that this is just version one of the product

Predictions About Sora's Growth and the Shift to Multiplayer AI

Prediction of unprecedented download growth

Fastest app to 100 million downloads[4:08]
• The host predicts Sora will be the most downloaded app in history over a short period and the fastest to reach 100 million downloads once invite gating is removed
Current invite gating inconsistency[4:29]
• He notes some users, including himself and the co-host, could download Sora directly from the app store with no invite, while others hit an invite code wall

Concept of multiplayer AI vs single-player tools

Matt Mazzeo's framework for AI products[4:44]
• They reference a previous episode with Matt Mazzeo where he argued that most current AI tools (ChatGPT, image generators) are "single player" experiences
• Mazzeo expects the next big wave to be "multiplayer" AI apps where interactions with other people are core to the product, analogous to Facebook, Uber, Airbnb, and Slack
Sora as an early "AI-native" multiplayer app[5:20]
• The host describes using Sora with friends by generating videos of each other and trading them, which he found extremely engaging and funny
• He highlights the powerful pull of receiving a tag notification that someone made a video of you, likening it to an irresistible compulsion to check it immediately

Comparison to Facebook's photo tagging growth loop

Photo tagging as a viral mechanic[6:03]
• He recalls how Facebook's introduction of photo tagging turned a solo upload action into a social loop, where tagged people received emails and were pulled into the platform
• He believes Sora's tagging and social sharing of AI videos will create a similar "insane growth loop"

AI Super Apps: WeChat Comparison and ChatGPT Pulse

WeChat as a model for an "everything app"

Capabilities of WeChat in China[6:39]
• The host explains that WeChat allows users to chat, order groceries, send money, pay electricity bills, and buy furniture within a single app

ChatGPT Pulse as a personalized AI feed

Overview of Pulse[7:01]
• The co-host describes subscribing to ChatGPT's most expensive tier to access a feature called Pulse, which functions as a daily, highly personalized feed
• Pulse learns the user's interests from explicit input and past conversations and presents AI-written articles and content tailored to those topics each morning
Pulse as both platform and content creator[7:23]
• He likens Pulse to Apple's News Feed except "it's also the journalist," since AI is both curating and generating the content
• Examples of interests he feeds into Pulse include a specific book he is reading, Greek philosophy, and strategies for YouTube storytelling and growth
• Pulse writes daily articles based on recent podcast-related research conversations they have had with ChatGPT, effectively pre-digesting information for him

AI as a "super app" and concerns about OpenAI's dominance

Comparison of Pulse to a human curator friend[7:23]
• One host compares Pulse to their friend Ben Levy, who is known for understanding people's goals and regularly texting them relevant links and encouragement
• He notes that Ben now has to "compete with AI" because AI is doing that same high-touch curatorial role at scale
Fear of an all-powerful AI company[8:46]
• The host characterizes Pulse as the "first example of a super app" that makes him feel everything will be done inside ChatGPT going forward
• He self-identifies as a free market supporter but admits growing fear each quarter about OpenAI potentially becoming significantly bigger than any previous company and knowing more about him than any entity ever has
• He contrasts OpenAI's everyday ubiquity and data access with companies like NVIDIA, which, despite massive market cap, don't directly interact with his daily life or personal information
• He predicts OpenAI could effectively combine the roles of Facebook, Google, every news company, and several other major companies into one

Emotional Reactions to Rapid AI Progress and Content Automation

Feeling overwhelmed by constant AI breakthroughs

Sister's reaction to Sora[9:36]
• The host sent a Sora-generated video of himself to his sister, who replied that being his sister is tiring because he constantly surfaces new technologies she now "needs" to learn
• She expressed that she had just gotten comfortable in life and now feels pressure to get up and adapt again because of new AI tools
Exhaustion vs empowerment[9:55]
• The host admits he also feels exhausted when hearing about yet another game-changing AI feature like Pulse, questioning whether he is now "super powerful or irrelevant"
• He uses a Survivor TV show analogy: contestants complain about being cold and wet, but the only reality is they must endure many more days in the rain, implying constant change is unavoidable

AI-generated podcasts and content creator relevance

Steve Bartlett's AI podcast experiments[10:59]
• They reference Steve Bartlett, host of "Diary of a CEO," saying "you're either the hunter or the hunted" regarding disruption in his own media business
• Steve reportedly has a team working on AI versions of himself and AI-generated podcasts, testing them with ad spend to compare watch times against his real podcast
• He has observed that average watch time on AI podcasts has reached parity with human-hosted episodes, which unsettles the hosts
Questioning human creators' future role[11:44]
• The host contemplates that as AI makes content creation extremely easy, human content creators may become irrelevant or alternatively become 10x more powerful using AI tools

Balancing Dystopian Fears with Optimism About AI

Metaphors of control and inequality

Palantir cage and haves vs have-nots[14:10]
• One host jokes that in the future, everyone will be in a "Palantir cage" while Sam Altman throws "slop" at them, symbolizing AI-driven control
• He predicts that AI will further separate the "haves and have-nots" in terms of power and capability

Three main emotional reactions to Sora and AI

Exhaustion at the rate of change[15:03]
• The host identifies one common reaction as exhaustion from constantly having to learn and adapt to new AI tools
Dystopian concern over power and misuse[16:23]
• He notes another reaction is dystopian fear: with Sora able to clone face and voice in seconds, anyone could generate videos of him doing or saying anything, including as an unpaid AI coach based on his writings and transcripts
• He points out that some people already distrust Sam Altman and worry about such powerful capabilities being centralized
Curiosity and play as the most useful stance[16:38]
• The third reaction he describes, and the one he tries to adopt, is curiosity and playful exploration of AI, viewing this as an amazing time to be alive
• He argues that being curious, having fun, and trying to positively influence the direction of change is the only useful reaction among the three

AI and the Future of Learning and Everyday Assistants

AI tutors and personalized education

Examples of AI tutoring initiatives[14:40]
• The host points to projects like Alpha School and a new Google feature called personalized learning that builds custom curricula based on what you say you want to learn
• He envisions users having not just therapists or sponsors in their pockets, but also a "Socrates in your pocket" for learning

Speculation on future AR-style AI assistants

Naming and behavior of future assistants[15:07]
• They joke about how big tech companies give bland names like "Maps," and imagine a more personable assistant named "Ava" who is "your cute friend who's all-knowing"
• They riff on use cases like asking Ava aloud in public how much a house is worth or whether someone has an OnlyFans, pointing to potential creepiness with AR glasses and always-on recording

Sports Micro-Betting, Prediction Markets, and Gambling Risks

Rise of constant micro-betting during sports games

Friend's habit of inning-by-inning betting[16:58]
• One host describes watching a baseball game with someone who placed $3-$5 bets between each inning via an app that uses AI to create micro-bets on what will happen next
• When he asked others in the room, nearly every man either did similar sports betting or knew a male close to them who did, suggesting it's widespread
Integration of gambling into sports culture[18:04]
• They note that virtually every ad during sports broadcasts and many podcast sponsors are companies like FanDuel and DraftKings
• Professional leagues that once banned gambling now embrace betting companies as premium sponsors

Impacts on players and integrity concerns

Examples of betting-related scrutiny[18:09]
• They mention several recent cases where NBA and NFL players were investigated due to suspicious plays that affected betting outcomes, such as not taking certain shots
• One specific incident involved a quarterback deliberately running backwards into his own end zone to drain the clock, turning a seven-point win into a five-point margin that flipped the betting line
• While that quarterback was not cheating, he and others now receive daily death threats from gamblers complaining that they "messed up my parlay" or over-under bets

Sports betting vs prediction markets and legal loopholes

Diverging fortunes of Robinhood and DraftKings[19:44]
• They reference a tweet showing Robinhood's stock price trending downward while DraftKings trended upward over the same period
How prediction markets enable betting features[20:14]
• The host explains that traditional sports betting means wagering directly on game outcomes and is tightly regulated by state gaming commissions
• Prediction markets, by contrast, are treated as financial markets where users buy and sell "shares" of outcomes, categorized more like commodities or futures
• Apps like Robinhood, Kalshi, and Polymarket now offer sports-related instruments via prediction markets, even in jurisdictions where conventional sports betting is restricted
• He notes that trading volume on these platforms is "exploding" after adding sports-style wagers under the prediction market label

Gambling Addiction, New Treatment Startups, and AI Sponsors

Concerns about ubiquitous, low-friction gambling

Distinction between casino nights and app-based habits[22:30]
• One host dislikes gambling but sees a difference between an occasional casino outing and constant, isolated micro-bets on a phone with promotional credits
• He mentions that some platforms assign account managers who call users offering large credits (e.g., $500) to incentivize more betting

Birches Health: telehealth for gambling addiction

Company overview and funding[23:05]
• He discusses Birches Health (bircheshealth.com), a company that raised $20 million from General Catalyst and Kevin Ryan, focused on helping especially young men with gambling addiction
Market size and growth signals[23:32]
• He cites a New York Times article stating that 60% of men aged 18-22 engage in sports betting and that a large percentage are now problematic gamblers
• Birches Health, about two years old, reportedly grew 5x in the last year, suggesting rapidly increasing demand for gambling addiction support
• He frames this as analogous to searching years earlier for alternatives to Alcoholics Anonymous as a big, impactful business opportunity

Previous ideas about physical centers for gambling addiction

Repurposing declining retail for treatment centers[24:47]
• They recall a prior podcast guest proposing repurposing dying retail concepts (like a Blockbuster-type chain) into physical centers for gambling addiction treatment

Alcoholics Anonymous, Sunflower Sober, and AI-Based Sponsors

Overview of Sunflower Sober app

AI sponsor concept[24:51]
• They introduce Sunflower Sober (sunflowersober.com), an app created by a listener named Kobe that serves as an AI-based version of a sobriety sponsor
• The idea is to have an always-on, nonjudgmental companion in your pocket to talk to when dealing with addiction challenges

Personal experience with Alcoholics Anonymous

Structure and culture of AA meetings[25:42]
• The co-host shares that he previously attended AA because he lacked money for rehab but needed help
• He describes AA meetings of 20-30 people sitting in a circle, many smoking cigarettes and drinking black coffee in the evening
• Participants go around saying their name and that they are an alcoholic, optionally sharing their story and how long they have been sober
Role of sponsors in AA[26:24]
• After some time, members get a sponsor they can call when tempted to drink or after a relapse, and sponsors are expected to go "above and beyond" to support them
• He emphasizes that AA and its sponsor system are free and built on people helping each other
Origins of the phrase "friend of Bill"[27:09]
• He explains the semi-secret phrase "Are you a friend of Bill?" which signals AA membership, referencing founder Bill W.
• Bill W., who founded AA in 1935, created a 12-step program and reportedly experimented with LSD as part of his journey, described as a "cowboy" story

Sunflower Sober's growth and business model possibilities

User and revenue growth[28:31]
• Sunflower Sober grew from zero to about 100,000 users in six months and reached over $50,000 in monthly recurring revenue from subscriptions
Potential expansion beyond alcohol[28:21]
• While initially motivated by a family member's alcoholism, the founder plans to extend the app to other addictions such as gambling and potentially more
AI coaching vs human therapy[29:03]
• They note that AI sponsors and therapists may be better in some ways (nonjudgmental, always available, infinitely patient) but possibly weaker in accountability compared to humans
Lead generation and vertical integration opportunities[29:29]
• Sunflower Sober passes hundreds of leads per month to teletherapy clinics for users who seek more clinical solutions, including medication-assisted treatment
• They speculate that the company could vertically integrate by creating its own clinic, dramatically increasing the lifetime value of each customer

Shifting Addiction Patterns and AI as Therapist/Coach

Decline of alcohol use among young men

Generational change in drinking culture[30:25]
• They note that alcohol consumption among young men has "plummeted" compared to their college years, when drinking multiple nights a week was common
• One co-host references "21 Jump Street" where older characters discover that bullying and homophobic jokes are no longer socially acceptable, paralleling changing views on heavy drinking

Comparison to meditation apps Calm and Headspace

Addiction recovery as the next big wellness vertical[31:50]
• The host compares Sunflower Sober's potential to Calm and Headspace, which turned meditation into a daily in-app ritual
• He predicts that helping people quit addictions like porn, gambling, alcohol, weed, and social media will become a huge opportunity, with apps making it a regular practice

Using ChatGPT as an executive coach and informal therapist

Applications for emotional processing and conflict[32:03]
• One host describes using ChatGPT to analyze his anger toward his wife, asking the AI to help him understand why he is upset and who might be in the wrong
• He also uses ChatGPT to draft measured responses to upsetting emails, aiming to avoid negative reactions while still achieving his goals
• He says ChatGPT has helped him "tremendously" in these areas, while the other host uses AI more for executive coaching-style habit tracking than therapy

AI expanding global access to mental health support

Cost, access, and stigma barriers[33:57]
• They argue AI can deliver services previously costing $100+ per hour for as little as $9 per month, making support accessible to more people worldwide
• AI may help people in places where therapists or AA meetings are rare, and reduce embarrassment that prevents some from seeking human help

Life-Logging, Personal Data, and Memory Capture

Tobi Lütke's keystroke and screenshot archive

Description of the logging system[34:54]
• They discuss an interview on the Acquired podcast in which Shopify CEO Tobi Lütke casually mentioned building a program 15-20 years ago that logs every letter he types
• The system also takes a screenshot of his screen every 10 minutes, resulting in roughly 15 years of complete keystroke and visual activity archives
Comparison to Marcus Aurelius's "Meditations"[36:07]
• They compare this archive to Marcus Aurelius's private diary, which unintentionally documented the inner thoughts of the most powerful man in the Roman Empire
• They argue that Tobi's dataset may be even more comprehensive because it bypasses the filter of writing and captures every distraction, tab, and action
• One host suggests this gives Tobi an unprecedented dataset to train an AI that thinks like him, with a full record of his founding years at Shopify

Personal practices for archiving memories

Home videos and "secret YouTube"[39:03]
• One host keeps a private YouTube channel where he uploads 5-10 minute unedited home videos each day capturing everyday family life
• He prefers longer, candid recordings that feel like old camcorder tapes, rather than the typical 10-second clips people capture on phones
Inspiration from the "One Second a Day" app[39:51]
• He used an app that compiles one-to-three-second daily clips; during his daughter's first year he recorded a short video every day and assembled them into a montage that makes him cry when he watches it
• Seeing his deceased dog and family moments in that montage underscores the emotional power of longitudinal memory capture
Desire for behavioral time tracking[40:42]
• He recalls using the RescueTime app, which tracked how he spent time on his computer, and wishes it could have told him even more about his behavior
• Tobi's keystroke logging system appeals to him as a way to look back years later at his evolution, not for productivity optimization but for self-understanding and narrative

Different philosophies on tracking vs memories

Tracking for behavior change vs over-tracking[41:18]
• The other host only tracks when trying to change behavior, such as food, weight, exercise, and whether he wrote for the first two hours of the day
• He views generalized tracking as "nervous energy" and believes some things, like sleep quality, are obvious without metrics-"How did you feel when you woke up?"
• He acknowledges, however, that devices like the Oura Ring are popular, indicating many people enjoy or feel reassured by detailed self-metrics
Archiving for memory vs optimization[41:56]
• The first host clarifies that his interest lies in memory archiving-like saving his childhood clothes for his daughter-rather than minute-by-minute optimization
• He reiterates fascination with Tobi's system as a tool for long-term self-reflection and personal history

Meta Smart Glasses as a Tool for Capturing Life

Current capabilities of Meta glasses

Hands-free video capture from POV[43:02]
• The host praises Meta glasses as "awesome" for recording two-minute clips from the wearer's point of view with a simple button press
• He notes they are particularly useful for parents at kids' sports events, letting them both watch and record without fumbling with a phone
Audio features and advantages over phones[44:15]
• In addition to video, the glasses function as discreet headphones with audio only the wearer hears, reducing the need for AirPods
• He emphasizes that taking out a phone often pulls people out of the moment and prompts others to "perform," whereas glasses allow more natural capture

Concerns about future versions adding screens and feeds

From pure capture device to attention sink[44:13]
• He anticipates Meta will make future versions "good bad" by adding screens that display feeds like Instagram or email while walking around
• He finds the current camera-only version wise because it focuses on capturing moments without becoming another distraction

Commitment to adopt and cost trends

Decision to buy[44:59]
• The co-host says this is the third time the glasses have been recommended and agrees to buy them, especially for capturing his child's moments
Price evolution[45:09]
• The recommending host notes that the glasses used to cost around $500 but are now "way cheaper" and improving with each iteration

Closing Tease About Another AI Tool

Future discussion on an empowering AI tool

Cliffhanger about a different AI product[45:14]
• The host mentions he has been experimenting with another AI tool that made him feel profoundly empowered and plans to lead with it in the next episode

Lessons Learned

Actionable insights and wisdom you can apply to your business, career, and personal life.

1

When confronted with rapid technological change, adopting a stance of curiosity and playful experimentation is more productive than exhaustion or dystopian paralysis.

Reflection Questions:

  • • What new technology have I been instinctively resisting that I could instead approach as a sandbox to learn and play in?
  • • How might reframing my fear of becoming irrelevant into an experiment in becoming more powerful with new tools change my daily habits?
  • • What is one small, low-risk way I can deliberately try a new AI tool this week just to understand its potential and limitations?
2

The most powerful digital products tend to be "multiplayer"-they create social loops (like tagging and sharing) that pull others in, rather than existing as isolated single-player tools.

Reflection Questions:

  • • Where in my current product, service, or work could I add a genuinely social interaction that makes it more valuable when others are involved?
  • • How could thinking in terms of growth loops (e.g., notifications, tagging, sharing) improve the way I design or distribute what I create?
  • • What is one existing tool I use that is still "single player," and how might I plug it into a community or team workflow to multiply its impact?
3

AI can dramatically expand access to coaching and therapy-like support by offering always-on, low-cost, nonjudgmental conversations for processing emotions, habits, and decisions.

Reflection Questions:

  • • In what situations do I currently wish I had a neutral coach or therapist on demand to help me think through my reactions or choices?
  • • How could I start using an AI chat tool as a first pass for untangling difficult emotions or drafting better responses before I act?
  • • What boundaries or guidelines do I want to set for myself so that I use AI support thoughtfully without avoiding important human conversations?
4

Frictionless digital gambling and micro-betting can quietly become pervasive, so it is important to recognize how design and incentives may be nudging people toward addictive behaviors.

Reflection Questions:

  • • Where in my own digital life do small, frequent, low-stakes actions (bets, purchases, likes, refreshes) add up to a habit I'm not fully in control of?
  • • If I design products or experiences, how might I unintentionally be encouraging compulsive use, and what safeguards could I build in?
  • • What clear personal rule could I set about money and risk (e.g., limits, time windows, or contexts) to keep entertainment from drifting into addiction?
5

Systematically capturing your life-through logs, home videos, or other archives-can create a powerful record for reflection and learning, but it works best when aligned with a clear purpose rather than tracking everything by default.

Reflection Questions:

  • • What aspects of my life or work would be most meaningful to look back on 10-15 years from now, and how could I start capturing them today?
  • • How might I separate tracking for behavior change (e.g., habits, writing) from archiving for memory (e.g., family moments) so I don't overwhelm myself?
  • • What is one simple logging or recording ritual I could start (daily or weekly) that would give my future self a richer, more honest picture of who I am now?

Episode Summary - Notes by River

The App That'll Be Bigger Than TikTok
0:00 0:00