Key insights

• 💼 Transitioning to new AI-driven roles after working at Instagram
• 🧠 Challenges of working with emerging AI capabilities
• 🏢 Differences in working with Enterprise customers
• 🔄 Adapting instincts to new product development processes
• 🔍 Navigating uncertainty and discovery within the company
• 🔬 Partnering with research for product innovation
• 📈 Designing for model's success rate
• 🌎 Real-world evaluation challenges
• 🧪 Models not being intelligence-limited but eval-limited
• 🎯 Understanding the problem and defining success
• 🔄 Automating processes and iteratively improving AI features
• 📊 Evaluations as a core skill for PMs
• 🤖 Prototyping with AI as an underused skill
• 🔧 Going deeper into the tech stack as an important skill for PMs
• ⚙️ Adapting to the non-deterministic nature of AI products
• 📚 Educating end users at scale
• 🚀 Fast adaptation to new technologies
• 🎓 Letting the product be educational
• 📢 Challenges and strategies in educating entire organizations about AI products
• 🍕 AI becoming valuable for non-experts
• 🤖 Use cases: ordering pizza, automating repetitive tasks, orchestrating multiple models for complex tasks
• 🔮 Future of AI: proactive models, sophisticated reasoning capabilities
• 🗣️ Proactive and asynchronous communication is key for the future of AI
• 🧠 Models are getting smarter at an accelerating rate
• 🎙️ Advanced interaction capabilities, such as voice mode and translation, are becoming more common
• 😊 AI's personality and empathy shape user interactions and relationships
• 💭 AI's ability to customize its personality raises interesting questions

Q&A

• What does the future of AI involve in terms of communication and interaction capabilities?

  The future of AI involves proactive and asynchronous communication, smarter models, advanced interaction capabilities like voice mode and translation, and AI's personality and empathy playing a crucial role in user interactions and relationships.

• What are some valuable use cases of AI for non-experts?

  AI is becoming valuable for non-experts with use cases including ordering pizza, automating repetitive tasks, and orchestrating multiple models for complex tasks. The future of AI involves proactive models and more sophisticated reasoning capabilities.

• How do AI professionals adapt to the non-deterministic nature of AI products?

  AI professionals adapt to the non-deterministic nature of AI products by educating end users at scale, letting the product be educational, and strategizing ways to educate entire organizations about AI products.

• What skills are essential for success as a product manager in AI-driven organizations?

  Essential skills for success as a product manager include understanding the problem and defining success, automating processes, writing evaluations, and iteratively improving AI features. Evaluations are becoming a core skill, and prototyping with AI and going deeper into the tech stack are vital for product managers.

• How do AI professionals navigate uncertainty and discovery within a company?

  AI professionals navigate uncertainty and discovery by partnering with research for product innovation, designing for model's success rate, and addressing real-world evaluation challenges.

• What are the challenges of transitioning to new AI-driven roles after working at Instagram?

  Transitioning to new AI-driven roles involves adapting instincts to new product development processes, navigating uncertainty and discovery within a company, and fast adaptation to new technologies.

• 00:18 Two former Instagram executives share their experiences transitioning to new roles in AI-driven organizations and the challenges of working with emerging AI capabilities and Enterprise customers.
• 07:20 Navigating uncertainty and discovery within a company, partnering with research for product innovation, designing for model's success rate, real-world evaluation challenges, models not being intelligence-limited but eval-limited
• 13:47 Success as a product manager involves understanding the problem, automating processes, writing evaluations, and iteratively improving AI features. Evaluations are becoming a core skill for PMs. Prototyping with AI and going deeper into the tech stack are essential skills for product people.
• 20:24 Adapting to the non-deterministic nature of AI products, educating end users at scale, fast adaptation to new technologies, letting the product be educational, challenges and strategies in educating entire organizations about AI products
• 26:58 AI is becoming more valuable for non-experts, with use cases including ordering pizza, automating repetitive tasks, and orchestrating multiple models for complex tasks. The future of AI involves proactive models and more sophisticated reasoning capabilities.
• 33:56 The future of AI involves proactive and asynchronous communication, smarter models, and advanced interaction capabilities. The AI's personality and empathy play a crucial role in user interactions.