Lindy Teammate: Flo Crivello on Multiplayer Agents, Memory & Why He’d Ban the Chinese Models He Uses

Lindy Teammate: Flo Crivello on Multiplayer Agents, Memory & Why He’d Ban the Chinese Models He Uses

More

Descriptions:

Nathan Labenz of the Cognitive Revolution interviews Flo Crivello, founder and CEO of Lindy, on the occasion of launching Lindy Teammate — an AI employee that lives in Slack, connects to company tools, and accumulates shared team context. Crivello frames the shift as a move from “single-player” to “multiplayer” AI, drawing an analogy to the difference between emailing document revisions and Google Docs: if you want AI to be a real teammate, it needs to be where the team collaborates, with its own persistent shared memory. The launch directly competes with Claude-based products and runs on DeepSeek for cost reasons — a tension Crivello addresses candidly given his stated position that Chinese models should be banned in the United States.

The first half of the conversation is a detailed technical deep-dive: how Lindy implements memory (including background processing loops that continually optimize what gets remembered), the social contracts around historical data access in shared contexts, and vendor recommendations for infrastructure primitives. Crivello endorses E2B for sandboxing and BrowserBase for browser management, while noting that Lindy built its own eval and observability platform in 2022 before commercial options existed — an investment he would avoid today.

The conversation closes with bigger-picture views: Crivello’s belief that we are in a “centaur era” of human-AI co-creation that will be short-lived, his assessment that AI is in takeoff as of mid-2026 without resolved alignment, and a nuanced debate with Labenz over whether a Chinese model ban is the right policy response versus something like mandatory insurance requirements for AI service providers.


📺 Source: Cognitive Revolution “How AI Changes Everything” · Published August 10, 2026
🏷️ Format: Podcast