https://video.ut0pia.org/videos/watch/77c38f72-532c-4319-8e21-ca136e909bf2 A compromised release of litellm, a Python package pulling three and a half million downloads a day, sat live for three hours installing a credential harvester for API keys, SSH keys and crypto keys along with a backdoor for remote command execution. It was caught by pure luck: the malware had a bug that crashed Cursor, and a researcher went looking. Saoud Rizwan uses it to mark how far the trust model has fallen. Zig's code of conduct now bans AI from pull requests, issues and even comments, because the team values growing contributors over collecting contributions. curl is weighing the end of a decades old bug bounty as AI generated reports drown it. tldraw closes pull requests on sight. GitHub shipped a switch to disable third party pull requests altogether, which is the feature that made GitHub what it is. He built Cline in the open, and his claim is that the community half of open source is the half that died. The half that survives is open weights, and the case is economic. Cline pitted GLM against Opus on a real bug in their own repository: GLM spent twice the tokens at half the cost, cleaned up dead code and confirmed the build compiled, while Opus finished faster with fewer tool calls but left type errors that broke the production build. Coinbase defaulted its internal gateway to GLM and Kimi and cut AI spend nearly in half. The precedent Rizwan reaches for is Open Compute: Facebook gave away its data center designs, the supply chain standardized on them, manufacturers moved to huge uniform production runs, and the commoditization that followed drove Facebook's own costs down by billions. His closing ask goes to the American labs. Release open weights models, because once the industry standardizes on foreign ones, marginal quality gains may not be enough to bring anyone back. Speaker info: https://x.com/sdrzn, https://github.com/saoudrizwan, https://cline.bot, Timestamps 0:00 Introduction to Cline and the early open source coding agent era. 1:30 The decline of the open source community and the rise of AI-driven distrust. 2:22 How projects like Zig, curl, and tldraw are responding to AI-generated noise. 3:45 Systemic risks: The litellm supply chain compromise example. 5:22 The economic case for the survival of "open weights" models. 5:49 Real-world impact: Corporate AI spending and infrastructure lock-in. 8:50 Comparing model intelligence vs. system-level AI verification (GLM vs. Opus). 10:57 The Open Compute precedent: How open standards commoditize the industry. 12:36 Future projections for AI inference costs and hardware capacity. 14:22 A call to action for American labs regarding open weights models. 16:01 Cline's shift to an open weights subscription model. Quotes (1:48) "GitHub is effectively an archive of SLOP PRs and issues and security reports." (3:47) "When I say that open source is dead, I mean some parts of it—like the community—it's just not worth cultivating anymore." (7:00) "They're essentially going to subsidize this until they have as many engineers dependent on their tooling as possible... and then inevitably the price gouging." (14:35) "Before we know it, all this infrastructure that we're investing in could be built on foreign models that take the world by a storm." Sat, 08 Aug 2026 17:26:12 GMT https://validator.w3.org/feed/docs/rss2.html PeerTube - https://video.ut0pia.org https://video.ut0pia.org/lazy-static/avatars/0287a09a-aae7-4840-9843-b416426e7046.webp https://video.ut0pia.org/videos/watch/77c38f72-532c-4319-8e21-ca136e909bf2 All rights reserved, unless otherwise specified in the terms specified at https://video.ut0pia.org/about and potential licenses granted by each content's rightholder.