AI agents will replace 80% of apps - OpenClaw creator explains | Peter Steinberger and Lex Fridman
Quick Overview
Peter Steinberger, the creator of gogcli, explains that AI agents will likely replace up to 80% of existing applications by providing a unified command-line interface (CLI) for Google services, leading to a shift where companies like Google might resist direct API access due to competitive concerns, but users will still find ways to interact with services.
Key Points: Steinberger predicts AI agents will make approximately 80% of current apps obsolete by providing a unified interface for Google services like Gmail, Calendar, and Docs. The gogcli tool is a script-friendly CLI designed to manage multiple Google Workspace services from the terminal, offering features like searching Gmail threads, managing calendar events, and handling Drive files. A major challenge discussed is that large companies, especially Google, may push back against direct API access for AI agents to prevent them from easily extracting data, potentially leading to competitive resistance. If companies make it difficult to access data via API (e.g., requiring explicit user interaction for every action), the agent must adapt, potentially by copying text or using simulated web interaction. The speaker suggests that in the future, companies might be forced to offer a unified API or risk being bypassed, as users will gravitate toward seamless solutions that work across services. The concept of an agent that knows the user's context (like sleep quality influencing workout planning) represents a superior level of utility compared to current standalone apps. The speaker uses the example of Uber's ordering process, noting that if an agent could handle the entire process, the user experience would be much better than current app-based interactions.
Context: This segment features Lex Fridman interviewing Peter Steinberger, the creator of gogcli (Google in your terminal), a command-line interface tool for interacting with various Google services. The discussion centers on the transformative potential of AI agents in software development and how they might disrupt the traditional app ecosystem by offering consolidated, context-aware functionality across multiple platforms, contrasting this with the current reliance on individual, often siloed, applications.