By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.
My First Million9 Business Ideas To Profit From The 2022 A.I. Gold Rush (Using Dall-E & GPT3) (#376)
Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.
Great conversations run long. Their best ideas don't have to.
uListen is free on iPhone, iPad, Android, Chrome, and Claude.
See howCHAPTERS
Jump to publisher timestampsChapter summaries
Feeling the new era: awe, fear, and “how do we get our paws on this?”
Sam says he feels like there’s a new era he’s not part of, and he’s both excited and unsettled by what he’s seeing in AI. Shaan frames the episode as an all-AI deep dive and describes it as his biggest “holy shit” technology moment.
- •Sam compares his reaction to seeing a new era emerge and feeling left behind.
- •Shaan sets the premise: this episode is entirely focused on AI and its implications.
- •They position themselves as opportunistic builders/investors trying to understand where the leverage is.
AI-generated podcast demo: Joe Rogan interviews Steve Jobs (Podcast.ai / Play.ht)
Sam describes a project where an AI model uses Steve Jobs recordings and biographies, plus Joe Rogan episodes, to generate a new interview between them. They play the opening moments and react to how convincingly it mimics the show format and voices.
- •Sam explains how the project was created using biographies and large audio back catalogs.
- •They play the generated intro and note how it imitates Rogan’s style and pacing.
- •They react to the realism: the voices, the banter, and the “showmanship” of the introduction.
Why it hits emotionally: realism, philosophy, and the “not really him” discomfort
Shaan and Sam break down why the generated conversation feels so real, including philosophical turns that don’t sound like a simple fact recap. Sam says the experience made him emotional and scared, especially when he found himself persuaded by lines attributed to Jobs.
- •Shaan highlights voice quality, conversational structure, and natural back-and-forth as the shock factor.
- •Sam describes feeling both awe and fear, especially when he agrees with what the AI “Jobs” says.
- •They discuss a segment about LSD and spirituality; Shaan notes at least one quote is real and woven into the conversation.
- •Ben adds that it’s compelling but frightening, imagining this being done with a deceased family member and the emotional consequences.
“Ask Aristotle a question”: the future gets weirder (and interactive)
Sam shares a 1985 Steve Jobs quote about capturing someone’s worldview in a computer so students could ask questions and get answers. They then discuss how Podcast.ai lets people vote on hypothetical conversations, and why even “make-believe” dialogue can feel powerful.
- •Sam quotes Jobs imagining interactive conversations with historical thinkers captured in software.
- •Shaan reacts that Jobs’ prediction resembles what was just done to Jobs himself.
- •Sam lists user-voted episode concepts and argues that simulated conversations can still help people explore perspectives.
- •They note the line between entertainment, insight, and manipulation is getting thin.
Shaan’s framework: the shift from “left brain” AI to “right brain” creative AI
Shaan explains a big-picture model: the last decade centered on analytical, decision-making AI, while the current wave is creative generation across text, images, audio, and video. He argues this is pulling in a broader audience—beyond “nerds”—because it impacts artists and creators directly.
- •Shaan contrasts analytical AI (chess, Go, self-driving decisions) with creative AI (writing, art, audio).
- •He points to GPT-3-like text generation and Dall-E-like image generation as the creative inflection point.
- •He notes audio generation (podcasts/voices) and video generation as the next layer.
- •He compares this moment to NFTs expanding crypto interest, but says this wave is bigger.
Runway ML demo: text-driven video editing and the “it’s not there yet, but it will be” moment
They react to a Runway ML demonstration where video scenes can be manipulated via simple instructions—removing objects, changing style, swapping backgrounds, and more. Both emphasize that even if current demos overpromise, the direction is obvious and fast-moving.
- •Shaan describes editing video by prompting changes like adding rain, removing objects, or changing visual style.
- •They react to background removal, green-screen-like effects, and fast scene transformations.
- •Sam compares this to early Tesla moments where limitations were clear but the trajectory was undeniable.
- •They use “puberty” as an analogy: imperfect now, but clearly developing toward maturity.
Business process disruption: sales emails, call centers, and “a robot that calls their robot”
Sam predicts automation of high-ticket sales back-and-forth and call-center style work as AI becomes capable of handling nuanced communication. Shaan shares an example from a company he invested in, where an AI system calls insurance phone trees to save medical offices time.
- •Sam describes the cost and training burden of sales email exchanges and argues automation is near-term plausible.
- •He predicts call centers shift toward AI-first operations with a small human fallback.
- •Shaan explains Infinitus’ approach: an automated agent navigates insurance verification systems with “infinite patience.”
- •They frame these as mundane but profound changes that cut real business friction.
Interior AI: instant re-imagining of rooms and removing creative friction
They discuss Peter Levels’ Interior AI tool, where users upload a room photo, pick the room type and style, and generate multiple redesigned versions instantly. They emphasize how this compresses the iterative loop of working with a human designer into button clicks.
- •Shaan explains the workflow: upload a photo, choose style, generate multiple renders.
- •They describe examples that dramatically transform spaces, sometimes in unrealistic ways.
- •Shaan argues it changes creativity: you can “dream out loud” and iterate instantly.
- •Sam reacts to a loft example where the tool decorates around structural features like pillars.
“If you’re great, you don’t have to be good”: imperfection today vs inevitability tomorrow
Shaan acknowledges the outputs aren’t perfect and points to a principle he attributes to a Paul Buchheit blog post: products can win by being great at a few key things while being weak elsewhere. They connect that to why these early AI tools are still so disruptive despite flaws.
- •Shaan points out current AI outputs can be weird or inconsistent, but improving fast matters more.
- •He cites the idea that being exceptional in a few core dimensions can outweigh many missing features.
- •He uses examples like critics focusing on what a product lacks rather than the breakthrough experience it enables.
- •They argue AI tools already remove enough friction to change workflows even before perfection.
How they plan to play the AI wave: investing, learning, and maybe building
Sam admits he doesn’t feel capable of building core AI tech and sees himself as more of a spectator or partial participant. Shaan says his strategy is to invest broadly and help technical founders with pain points and go-to-market, while staying open to starting something himself.
- •Sam describes feeling outmatched and compares it to seeing new content eras like TikTok.
- •Shaan argues many builders don’t know where to apply the tech or how to make a defensible business.
- •Shaan’s plan: invest in many companies, keep learning, and possibly start a company later.
- •Sam’s extended office story becomes an analogy for how hard it is to ignore something attention-grabbing and disruptive.
Jasper AI: “Google Docs with an assistant” and the copywriting productivity leap
Shaan describes investing in Jasper and shows how it can generate sales emails, ads, and rewritten text on command, including via voice input. He shares how he’s experimenting with it for his writing course and newsletter workflows, while Sam compares it to other copy tools he’s invested in.
- •Shaan explains why Jasper’s output feels like a fast, creative employee who generates many options instantly.
- •They play a demo where a real estate agent produces a personalized outbound email via a spoken command.
- •Shaan describes using it to generate repeated newsletter elements like clever daily openers.
- •Sam contrasts Jasper with Copy.ai and questions differentiation; Shaan frames it as similar but stronger execution/revenue.
Defensibility and “business utility”: SEO scoring, product photography, and ad creative generation
Shaan argues many tools use similar underlying language models, so defensibility comes from workflow integration and practical features like SEO scoring. He then shifts to visual commerce: AI-assisted product photography and turning static inputs into ad-ready creative.
- •Shaan says the core models may be less defensible; the edge is in UI and embedded business workflows.
- •He describes AI-written blog posts plus built-in SEO scoring as an example of added utility.
- •He explains how e-commerce photography is expensive and slow, and why AI-generated imagery could replace much of it.
- •They react to examples where a basic hoodie selfie becomes a polished studio-style model shot and where products get “epic” backgrounds.
- •Shaan mentions Osmosis as a concept for turning images into ad creative and video-like assets.
The “real life Jarvis”: AI-driven game characters and the rise of AI friends
Shaan describes meeting young founders building an AI assistant vision inspired by Jarvis and an early product called Carter for gaming NPCs. Sam reads a prediction that children may have more AI friends than human friends, and they discuss loneliness as a major driver for adoption.
- •Shaan recounts founders meeting through TikTok while building “Jarvis” publicly.
- •He explains Carter’s pitch: plug-in conversational AI so in-game NPCs can talk dynamically and guide players.
- •Sam reads a claim that AI friends could become more common and more “reliable” than human friends.
- •They frame loneliness as a growing problem and suggest AI companionship could have meaningful upside.
Two startup ideas: AI therapy/companionship and “AI Spotify” that generates songs on demand
Shaan proposes an AI companion or AI therapist product as a massive opportunity, emphasizing lower cost and higher accessibility for people who need someone to talk to. He then pitches an “AI Spotify” concept: generated music that learns your tastes and can create ephemeral tracks that users could choose to keep or “mint.”
- •Shaan describes AI therapy/companionship as scalable, always-available, and potentially nonjudgmental.
- •He argues cost and access barriers in traditional therapy create a huge unmet market.
- •He critiques music platform economics and proposes AI-generated songs that avoid paying for traditional catalogs.
- •He imagines music generated on the fly based on your preferences and skipping behavior.
- •He adds a concept where a listener could choose to “own” a generated song by minting it, tied to their role in shaping it.
Billy of the Week: Emad Mostaque, Stable Diffusion, and the open-source AI mission
Shaan nominates Emad Mostaque and describes him as the force behind an open-source competitor to Dall-E. He recounts a personal and career story involving quantitative finance, family motivations, an interest in using AI for societal problems, and a push to keep powerful AI tools from being controlled only by private companies.
- •Shaan describes Stable Diffusion as a more open alternative to Dall-E and praises the “open” approach outperforming a closed one.
- •He recounts Emad’s background as a quant, stepping away after having a child, and becoming driven to apply AI to difficult problems.
- •Shaan says Emad adopted a “Promethean” mission to open-source powerful AI/ML tools broadly.
- •They discuss open-source business models and how monetization may come via usage, services, or related infrastructure.
- •Sam reacts to the boldness of the story and says it makes him feel inadequate by comparison.
Closing energy: inspiration, exhaustion, and “AI is all the way across the room”
They wrap up describing the episode as mentally exhausting in a good way—like being overstimulated by excitement. Sam jokes about needing a nap, and they end with a metaphor: AI feels important and compelling, but it’s still “all the way across the room,” requiring effort to fully commit.
- •Sam says the conversation leaves him hyped and drained, like he needs to rest afterward.
- •They joke about “dad chair” naps and the physical rituals of being exhausted.
- •Sam describes AI as a “quit and dedicate your life” topic—then admits he probably won’t.
- •They land on the metaphor that AI is meaningful but still requires crossing the room to engage deeply.
