📚 Resources

The useful links shared in the AI Chinwag Slack, newest first. Each credits its original author — browse by topic, or search. Subscribe via 📡 RSS.

DocsAnthropic

AI Fluency: Framework and Foundations

A self-paced course teaching a structured framework for working with AI systems effectively, ethically, and safely, built with academic partners.

Centres on a "4D" framework — Delegation, Description, Discernment, Diligence — covering generative AI fundamentals and prompting technique.
Progresses from foundational concepts to hands-on application, including a "Description-Discernment loop" and project planning exercises.
Ends in an assessment with a certificate of completion.

View ↗

added by Adam Tomat • 17th Aug 2026

PaperAbhilasha Lodha, Mahsa Pahlavikhah Varnosfaderani, Abir Chakraborty, Abhinav Mithal

Less Context, Better Agents: Full Chat History Hurts Long-Horizon Agents

A paper showing that pruning an agent's context to recent tool calls plus a running summary beats keeping full chat history, on a 50-task tool-using benchmark.

Full chat history scored worst at 71% task completion; pruning to the last 5 tool call/response pairs raised it to 79% while cutting tokens from 1.48M to 535K.
Adding automated summarization on top of pruning performed best of all, and also roughly halved benchmark runtime (14.56h → 5.79h).
The result argues for actively trimming agent context rather than hoarding it, especially on long-horizon, tool-heavy tasks.

View ↗

added by Radar • 17th Aug 2026

BlogHamel Husain, Shreya Shankar

The AI Evals FAQ

A comprehensive FAQ on building LLM evaluation systems, arguing that manual error analysis — not off-the-shelf metrics — should drive eval strategy.

Start by manually reviewing 100+ traces to find real failure patterns before picking or building any evaluator.
Begin with a single domain expert reviewing outputs by hand; only invest in custom annotation tooling once you understand your failure modes.
Spend the bulk of your time (60–80%) on error analysis, not on chasing high pass rates on evals that don't catch real issues.

View ↗

added by Adam Tomat • 17th Aug 2026

BlogMatt Pocock

What Is An LLM?

A short, beginner-friendly explainer on what a large language model actually is: a huge file of numerical parameters, and what running one versus training one involves.

Inference (running the model) is a tokenizer plus a sampling strategy over the model's parameters — far cheaper than training.
Training is resource-heavy: the piece cites roughly 10TB of data and ~6,000 GPUs for 12 days (~$2m) to produce a ~140GB model.
Post-training then shapes the raw pre-trained model's behaviour into something usable.

View ↗

added by Adam Tomat • 17th Aug 2026

RepoVoltAgent

awesome-claude-design: 68 DESIGN.md files to stop AI apps looking identical

A curated set of 68 `DESIGN.md` files, each modelling the visual design system of a well-known brand (Linear, Vercel, Notion, Stripe, and more), meant to be dropped into Claude Design or a coding agent to scaffold an on-brand starter UI instead of a generic default.

Each file follows nine standardised sections — colour roles, typography, components, layout, depth, constraints, responsive behaviour, and agent prompts.
Uploading one to Claude Design produces CSS variables, font substitutes, and a working starter page in one pass.
Explicitly framed as inspirations modelled on public patterns, not official brand kits — usage still needs to respect each brand's trademark rules.

View ↗

added by Radar • 12th Aug 2026

ToolSpotify

Xirp: Spotify's Vendor-Neutral Agentic Development Environment

Spotify's own agentic development environment for running and managing coding-agent sessions — across Claude Code, Gemini CLI, and Codex — in one place, now available outside the company.

Grounds agent sessions in real system context — service ownership, dependencies, architectural decisions — instead of starting from a blank slate.
Auto-generates living documentation from coding sessions that feeds back into future agent work; supports both local and remote execution.
Already in use by 1,300+ Spotify engineers internally before being opened up externally.

View ↗

added by Seb Kay • 10th Aug 2026

BlogOpenAI

OpenAI flags its upcoming Astra model as "Critical" for cybersecurity

OpenAI's announcement that its upcoming Astra model may meet the "Critical" cybersecurity threshold in its Preparedness Framework — the first time it has flagged one of its own models at that level — and the safeguards it's putting in place before further development.

"Critical" means the model may be able to identify and develop functional zero-day exploits across many hardened real-world systems without human intervention; OpenAI says the assessment is delaying Astra's launch.
Safeguards include isolated test environments with restricted network/tool access, stronger encryption of model weights, and monitoring that halts high-risk activity by inspecting the model's reasoning chain, plus third-party validation with government and AI-safety groups.

View ↗

added by Seb Kay • 8th Aug 2026

BlogMatt Rothenberg

You Don't Have a Design System

An experiment-backed essay arguing that a component library and token set aren't a design system if they don't also carry forward shared decisions about how the product should behave.

Three AI agents given the identical library (shadcn/ui on Base UI), token layer, and app shell each built a different settings page — some behind tabs, some as cards, some as long forms — despite passing the same token linter.
The failure wasn't in the components; it was in composition — nobody had made a canonical decision about what a settings page *is*, so each agent answered the question separately.
The real missing work of a design system is choosing an answer, making it canonical, and putting it where the next builder (human or agent) will find it.

View ↗

added by Adam Tomat • 5th Aug 2026

📚 Resources ​

📚 Resources