AI demos + Jason invests $25K live! | E1789

Episode Summary

Executive Summary: The episode focused on two themes: the disruptive effect of AI on content creation and productivity tools, and a live Founder University pitch session in which three early-stage AI startups each received a $25K investment. Jason and Sunny debated whether AI-generated content will improve workflows or flood the web with low-quality spam, while also arguing that vertical AI products will augment human performance rather than be replaced by base models.

Main Topics: AI content generation and web pollution (Priority: 5/5): Jason and Sunny reviewed AI tools that generate SEO pages and social clips, praising their speed while worrying they will create generic, spammy content that overwhelms the open web and search engines. Custom AI workflows and prompt control (Priority: 5/5): They demonstrated ChatGPT custom instructions and discussed how simple user preferences can materially improve output formatting, tone, and usefulness for specific tasks. AI as a productivity multiplier (Priority: 5/5): The conversation emphasized that AI will make small teams dramatically more productive, allowing two- or three-person startups to build products and content at unprecedented speed. Founder University and startup investing (Priority: 5/5): The back half of the episode featured three startup pitches from Founder University and culminated in Jason deciding to invest $25K in all three companies. Vertical AI startup opportunities (Priority: 4/5): The hosts argued that specialized AI products for analytics, productivity, and design will outperform general-purpose models because they embed workflow, integrations, and domain-specific value. Show strategy and audience fit (Priority: 3/5): Jason discussed evolving This Week in Startups’ format to better serve founders and VCs, balancing news with deeper founder interviews and recurring business breakdowns.

Key Arguments: AI-generated SEO and clip content can help with brainstorming and first drafts, but without human curation it risks becoming low-value filler that pollutes search and social feeds. Search engines and social platforms will need stronger detection and ranking systems to separate useful human content from automated output. Custom instructions in ChatGPT are a major usability improvement because they let users standardize tone, formatting, and output structure without re-prompting. The next wave of AI improvement will come from larger parameter counts and more capable models, not just more training data. Vertical AI tools will create more value than base LLMs alone because they combine model capabilities with integrations, domain context, and workflow design. The strongest early-stage teams are small, technical, and move quickly; AI can compress product timelines and make startups “bionic.” Founder-level clarity matters: Jason used pitch Q&A to evaluate not just the idea, but how founders think about users, market, and product strategy. Systematic guest and content planning improves a media business more than reactive booking or news chasing. Founder University is designed to meet founders early, help them launch, and identify companies worthy of later seed/accelerator investment.

Data Points: AI companies in market: 6,500 - Sunny cited an estimate for the number of AI companies currently operating. SEO page creation time: less than 5 minutes - Sunny built an optimized All-In Summit page using an AI content generator. LLM output improvement timeline: 6 months - Sunny predicted AI-generated content quality could reach near-human parity within six months. Open-source model size: 7 billion parameters - Sunny referenced open-source models runnable on a Mac as evidence of model scaling. Submagic clip editing time saved: 20-30 minutes per clip - Jason described how auto-captioning and clip generation can replace manual editing work. ChatGPT custom instructions: 2 text boxes - One box stores background info; the other controls response style and behavior. Founder University cohort size: 250 people - The program accepts up to 250 founders per cohort. Founder University applications: 1,500 applicants - Jason said roughly 1,500 people apply to each cohort. Founder University investments: 41 investments - Total investments made across all previous cohorts. Additional companies in accelerator: 11 - Companies that progressed from Founder University into the Launch Accelerator. Querio ARR: $12,000 annual recurring revenue - Rami reported early revenue from the data analytics startup. Before Sunset user growth: 10,000+ users - Batu said the productivity startup onboarded over ten thousand users after launch. Before Sunset launch traction: #1 Product of the Day/Week/Month - The team said it ranked first on Product Hunt across multiple time windows. Before Sunset market size: $5.8 billion - Founder estimated total market size for the productivity product. Symantic waitlist: 300+ - Alan said the design copilot had over 300 users on the waitlist. Symantic enterprise interest: 10 verbal commitments - Alan reported early enterprise demand. Symantic pricing: $45 / $85 / $120 per month - Per-editor pricing tiers for individual, team, and enterprise-oriented usage.

Pivotal Quotes: "I think this is going to flood and destroy the web if it's not." — Jason Calacanis: Reaction to AI batch content generation tools that can mass-produce SEO pages and clips. "What would it know about you to provide better responses?" — ChatGPT UI prompt label (quoted by Jason): Jason highlighted the new custom instructions feature in ChatGPT and its practical usefulness. "This is an AI native approach. It's totally rethinking how we build digital products." — Alan: Defense of Symantic as a design copilot competing with traditional UI tools.

Implications: AI will reward teams that combine models with workflow and taste, while punishing generic automation. For startups, vertical tools and prompt-driven personalization are immediate opportunities; for the web, spam filtering and human curation become more important.