A manifesto against low-quality, machine-generated output has gained traction in the AI community after being introduced at the AIE CODE conference. Dex presented the No AI Slop Manifesto in the event's closing keynote, after many of the speakers had added their cosignatures. The document, hosted at slopmanifesto.com with edits and discussion open to the public, quickly reached the front page of Hacker News. Its authors say they wrote it as a practical tool for teams to discuss how AI output should be used at work.
The article is part of Latent Space's daily AI news roundup, which also covers industry developments. The llama.cpp and ggml projects reached version 1.0, declaring API stability and removing 35 deprecated functions from llama.cpp's public header. The SGLang inference framework also announced 1.0. The llama.cpp release adds a cache for frequently used experts in mixture-of-experts models that are larger than a GPU's memory. Separately, LangChain raised a $300 million Series C at a $3 billion valuation, led by CapitalG, and plans to announce new LangSmith features for tracing and evaluating production agents on November 20.
For Australian businesses and professionals, the manifesto changes no legal obligations and carries no cost on its own. Its practical value is as a starting point for team discussion, such as agreeing review standards for AI-drafted documents, code and customer material before they go out. The local-inference releases are worth a trial for teams that want to keep data on their own hardware, which can simplify some Privacy Act questions around cross-border disclosure under APP 8, though self-hosting still requires security and governance work. The funding and model news does not change costs or obligations for most Australian teams this week.
Key Points
The No AI Slop Manifesto was introduced by Dex in the closing keynote of the AIE CODE conference and cosigned by many speakers.
The manifesto is hosted at slopmanifesto.com, where edits and discussion are open, and reached the Hacker News front page.
Its stated purpose is to give teams a discussion tool for how AI-generated work is used at the workplace.
llama.cpp and ggml reached 1.0, with SGLang also announcing 1.0, signalling more stable local inference tooling.
LangChain raised a $300 million Series C at a $3 billion valuation, led by CapitalG.