OpenAI Dev Day 2025: Apps SDK, Agent Kit, Sora 2, ChatGPT as operating system
In this episode of Generation AI, hosts Ardis Kadiu and Petar Djordjevic take you inside OpenAI's third annual Dev Day in San Francisco, breaking down the major announcements that are reshaping how we interact with AI. With ChatGPT now reaching 800 million weekly active users, OpenAI is positioning itself as the operating system of the future. Ardis and Petar, who attended the event in person, discuss three major announcement categories: Apps (native applications running directly in ChatGPT with deep integration), Agent Kit (a visual agent builder with built-in evaluation systems), and new models including GPT-5 Pro, Sora 2 video generation, and cheaper image options. They explore what these changes mean for developers, product builders, and higher education professionals, while sharing their first-hand observations from being in the room with 1,500 developers and AI industry leaders. This episode is essential listening for anyone trying to understand where AI platforms are headed and how to prepare for a future where ChatGPT becomes the hub for all your digital work.Dev Day Experience: San Francisco and the AI Ecosystem (00:00:36)First-time experience attending OpenAI Dev Day in San Francisco with 1,500 attendeesThe unique culture of San Francisco's tech scene and AI billboards everywhereMeeting AI influencers, builders from major companies like Netflix, Facebook, MicrosoftComparing Element451's AI work against world-class builders and feeling competitiveThe optimism and grind culture among new builders and startup foundersThe Three Big Announcement Categories (00:06:32)OpenAI's strategic shift: positioning ChatGPT as an operating systemThree main categories: Apps, Agents, and new ModelsChatGPT reaching 800 million weekly active users (not monthly - weekly)Processing billions of tokens daily across the platformApps in ChatGPT: The Third Try at an App Ecosystem (00:10:05)Native applications running directly in ChatGPT with deep integrationEvolution from plugins (first attempt) to custom GPTs (second attempt) to Apps SDK (third attempt)Launch partners: Canva, Booking.com, Expedia, Figma, Spotify, Khan Academy, Instacart, Uber, TripAdvisorApps can share context with ChatGPT and return custom UI componentsDemo showing Coursera courses, Canva slide creation, and Zillow apartment search all within ChatGPTApps SDK will be available to all developers by end of yearThe Distribution Flywheel and Vendor Lock-in (00:14:53)800 million users creates massive distribution leverage for app makersThe more users work inside ChatGPT, the more context gets centralizedThis strengthens personalization but also increases switching costsChatGPT becoming your memory and general assistantDiscussion of potential for ads and payment systems within ChatGPTUsers becoming more sticky to ChatGPT than to individual app websitesAgent Kit: Visual Agent Builder with Native Evals (00:18:38)Visual agent builder for orchestrating multi-agent workflowsChat Kit for embedding chat interfaces into applicationsNative evaluation system built directly into the platformLive demo: building a full agent for Dev Day conference in 8 minutes on stagePre-built guardrails for PII data and harmful contentConnections to file search, web search, and external systems via MCP protocolSimilar to tools like Zapier, Make.com, and n8n but with embeddable chat widgetsHow OpenAI Uses AI Internally (00:23:44)OpenAI shared three internal use cases at a breakout sessionGo-to-market agent: researches customers before meetings, preps demos, closes the loop after meetingsSupport agent: handles customer inquiries at scale (not outsourced, built in-house)When ChatGPT image generation launched, they got 10 million new users in a dayBuilt-in evals allow systems to improve themselves over time using thumbs up/down feedbackEvals and Prompt Optimization: The Game Changer (00:25:23)Evals explained: non-deterministic outputs require grading systemsEvolution from human graders to LLM gradersOpenAI introducing prompt optimization using the GEPA algorithm (Genetic Pareto)System uses all your data and feedback to automatically improve promptsConnection to DSPY library and the movement toward automated prompt engineeringNot locking users into OpenAI models - can use any model and send traces to the systemComparison with LangSmith and other tracing toolsNew Models: GPT-5 Pro, Sora 2, and Image Mini (00:33:20)GPT-5 Pro now available via API (12x more expensive than standard ChatGPT)Takes minimum 15 minutes per task due to deep reasoning capabilitiesSora 2 and Sora 2 Pro for video generation now in APISora app showing amazing video generation capabilitiesDemo with UK animation studio showing year-long process compressed to minutesGPT Image 1 Mini: 80% cheaper for cost-sensitive, high-frequency tasksEnables personalized images at scale for hundreds of thousands of usersTwo-tier Sora workflow: use smaller model to nail the prompt, then Pro for high fidelityReal-Time Voice Models and Device Strategy (00:40:38)GPT Real-Time Mini Voice: 70% cheaper with improved qualityDiscussion about voice quality expectations and production use casesSpeculation about OpenAI's strategy to get models small enough for on-device deploymentThe importance of voice as a natural interface for future applicationsConcerns about whether cheaper models sacrifice too much qualityCommunity Reactions and the Agent Debate (00:43:26)Mixed reactions to Agent Kit announcementsTwo camps: those excited about workflow builders vs. those disappointed it's "old paradigm"Debate about what defines an "agent" - no consensus in the industryComparison with Claude Code's different approach: treating LLM as autonomous humanDiscussion of workflow builders vs. true autonomous agentsWhat This Means for Startups and Builders (00:47:40)Advice: still build in code, don't rely entirely on Agent Kit for productionAgent Kit good for proof of concept and quick distributionWill take at least a year for App Store to catch fire with normal usersOpportunity to be early in the ChatGPT App ecosystemImportance of building expertise with OpenAI's tooling and platformThe Everything App and Multi-Platform Future (00:50:30)ChatGPT positioning as the "Everything App" and operating system of the futureGoogle announces Gemini Enterprise with similar agent builder capabilitiesQ4 2025 prediction: proliferation of agent builders across platformsElement451's approach: building agents that build agents using conversational interfaceEvolution from visual workflow canvas to AI-driven job creationProactive AI that evaluates context and takes actions without predefined stepsFinal Thoughts: The OpenAI Ecosystem (00:54:13)OpenAI as one of the most advanced AI labs with 4 million developers on platformChatGPT as dominant chat assistant with massive ecosystem impactKey takeaways from being there in person and seeing the builder communityHow these announcements will shape the future of work and higher education
- - - -Connect With Our Co-Hosts:Ardis Kadiuhttps://www.linkedin.com/in/ardis/https://twitter.com/ardisDr. JC Bonillahttps://www.linkedin.com/in/jcbonilla/https://twitter.com/jbonillxAbout The Enrollify Podcast Network:Generation AI is a part of the Enrollify Podcast Network. If you like this podcast, chances are you’ll like other Enrollify shows too! Enrollify is made possible by Element451 — The AI Workforce Platform for Higher Ed. Learn more at element451.com. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
New Zealand