Mike Krieger, co-founder of Instagram and lead of Anthropic Labs, discusses how the team bridges advanced AI model development with innovative product creation, exemplified by tools like Fable, Cloud Code, and Cowork, emphasizing AI’s enhanced capabilities in understanding complex tasks and contexts. He highlights Anthropic’s commitment to ethical AI, collaboration over competition, and future-focused innovations aimed at creating autonomous, environment-aware AI systems while advocating for outcome-based measures of AI productivity and value.
Mike Krieger, co-founder of Instagram and current lead of Anthropic Labs, discusses Anthropic’s dual role as both a major AI model builder and a product innovator. Anthropic Labs focuses on creating frontier AI-native products that push the boundaries of what AI can do, exemplified by successful offerings like Cloud Code and Cowork. Krieger explains that Labs was established to ensure that Anthropic’s products keep pace with the rapid advancements in AI models, bridging the gap between model capabilities and real-world applications. The Labs team experiments with new product forms and anticipates future AI improvements to prepare innovative solutions ahead of time.
Krieger highlights the significant impact of Anthropic’s advanced models like Fable, which, despite being temporarily restricted by the U.S. government, represent a leap in AI capability. He shares personal experiences of delegating complex tasks to these models, such as converting large codebases between programming languages, noting that the models are not only faster but also more accurate and reliable than previous iterations. This enhanced “theory of project” ability allows the AI to understand broader contexts and anticipate future challenges, fundamentally changing how developers and product builders interact with AI.
The conversation also touches on Anthropic’s approach to balancing its role as both a platform provider and a product creator. Krieger emphasizes transparency and collaboration with partners, aiming to push the industry forward rather than simply competing. He acknowledges the challenges startups face when large companies enter their markets but stresses that Anthropic’s goal is to innovate and create new spaces rather than crowd out others. This philosophy is reflected in their product strategy, which focuses on unique capabilities and shared building blocks rather than duplicating existing solutions.
Krieger reflects on Anthropic’s culture and mission-driven approach, contrasting it with his experience at Instagram. He praises Anthropic’s deep commitment to AI safety and ethical considerations, which permeates the company’s operations and decision-making. This focus on responsible AI development is seen as a defining characteristic that could influence Silicon Valley’s broader culture. He also notes the evolving conversation around AI’s societal impact, highlighting the importance of proactive engagement with ethical and safety issues as AI technologies rapidly advance.
Looking ahead, Krieger shares exciting directions for Anthropic Labs, including enhancing AI agents’ autonomy and environmental awareness to create more expressive and capable products. He envisions AI systems that can seamlessly interact with their environments and users, reducing friction and enabling complex problem-solving. While cautious about moonshot ideas like space-based data centers, he acknowledges their potential. Finally, Krieger discusses the challenges of measuring AI productivity and pricing, advocating for outcome-based approaches that better capture AI’s value beyond token usage, and underscores the ongoing effort to make AI tools more accessible and impactful across industries.
