Today's AI news: No AI Slop Manifesto, Jev models, CoReMem coding agents, open-source AI code debate, and Jensen Huang GPU shortage allegations.
The No AI Slop Manifesto: Engineers Push Back on Low-Quality AI Content
A coalition of AI Engineer (AIE) speakers has published what they are calling the No AI Slop Manifesto, a collective statement laying out concrete ways developers and creators can fight the proliferation of low-effort, AI-generated content — commonly dubbed "slop" — that has increasingly cluttered the web. The manifesto, surfaced via Latent Space's AINews newsletter, signals growing frustration inside the AI engineering community with the gap between what AI tools are capable of and how they are routinely used to churn out undifferentiated, low-value output. The document is framed as a call to action for practitioners to hold themselves and their tools to a higher standard.
The Jev Revolution: What Engineers Are Already Building with System One Models
At the AI Engineer Code Summit, Diogo Almeida — lead creator of Jev — called on developers to automate everything using what are being termed System One Models. The session highlighted how companies including Glean are already deploying Jev in production workflows. The Latent Space report frames this as a nascent but fast-moving shift in how engineers think about model deployment: rather than treating AI as a tool for discrete tasks, Jev-based approaches encourage comprehensive automation of entire engineering pipelines. The summit coverage suggests real-world adoption is further along than many outside the AI engineering world may realise.
CoReMem: Smarter Memory for AI Coding Agents
Researchers have introduced CoReMem (Coding-Repository Memory), a new framework designed to help AI coding agents work more efficiently across successive tasks on an evolving codebase. The core insight is that agents must decide what information is worth remembering before they know what the next task will be — a genuinely difficult planning problem. CoReMem addresses this through prospective segment valuation, which compares how well an agent performs on a subsequent task with and without specific pieces of remembered context, and builds compact memory around repository locations and executable checks called anchors. Tested on the ChainSWE benchmark, a 27-billion-parameter actor using CoReMem lifted its resolve rate from 21.0% to 25.7%, while taking 17% fewer steps and using 11% fewer total tokens compared to the same model without memory. For teams building coding agents at scale, those efficiency gains are meaningful.
Should All AI-Generated Code Be Open Source? The Debate Heats Up
A blog post arguing that all AI-generated code should be released as open source attracted 867 points on Hacker News, making it one of the most-discussed AI topics of the day. The piece, published by Jakub Jankiewicz, sparked a wide-ranging community debate about intellectual property, the provenance of training data, and the obligations that come with using models trained on publicly available code. While the article's specific arguments are not detailed in the available summary, the strong community response signals that questions around AI code ownership and licensing remain far from settled — and are becoming more urgent as AI-assisted development goes mainstream.
Jensen Huang Allegedly Financed Startups to Absorb Excess GPU Supply
A report from WCCFTech, which gained traction on Hacker News, alleges that NVIDIA chief Jensen Huang deliberately financed AI startups in order to absorb excess GPU silicon — effectively manufacturing the appearance of a supply shortage. The claim, if substantiated, would have significant implications for how the AI hardware market has been shaped over recent years. The story attracted 52 points on Hacker News, reflecting genuine public interest even if the evidence behind the allegation remains contested. NVIDIA has built much of its current dominance on sustained, high demand for its GPUs; any suggestion that demand was artificially stimulated would be a serious matter for regulators and investors alike.
DRAM Price-Fixing Case Survives Dismissal — Implications for AI Hardware Costs
A US court has denied a motion to dismiss a DRAM price-fixing lawsuit against Samsung, SK Hynix, and Micron, allowing the conspiracy case to proceed. The ruling, reported by LitNews AI and flagged on Hacker News with 444 points, is significant for the broader AI industry because DRAM is a critical component in AI accelerators and data centre infrastructure. If the plaintiffs ultimately succeed, it could reshape how memory pricing is scrutinised across the semiconductor supply chain — with downstream consequences for the cost of building and running AI systems.
Frequently Asked Questions
What is the No AI Slop Manifesto?
The No AI Slop Manifesto is a collective statement published by AI Engineer speakers laying out ways developers and creators can fight the proliferation of low-effort, low-quality AI-generated content. It was surfaced via the Latent Space AINews newsletter on 14 November 2026.
What is CoReMem and what results did it achieve?
CoReMem (Coding-Repository Memory) is a research framework that helps AI coding agents manage memory efficiently across successive tasks on an evolving codebase. On the ChainSWE benchmark, it raised a 27-billion-parameter model's resolve rate from 21.0% to 25.7%, with 17% fewer steps and 11% fewer total tokens compared to the same model without memory.
What are the allegations against Jensen Huang and NVIDIA regarding GPU supply?
A WCCFTech report alleges that NVIDIA chief Jensen Huang financed AI startups in order to absorb excess GPU silicon, effectively creating the appearance of a supply shortage. The claim has not been substantiated but attracted significant attention on Hacker News.
Why does the DRAM price-fixing lawsuit matter for AI?
A US court denied a motion to dismiss a price-fixing lawsuit against Samsung, SK Hynix, and Micron. DRAM is a critical component in AI accelerators and data centre infrastructure, so the outcome of this case could affect memory pricing across the semiconductor supply chain and the broader cost of building AI systems.
