Slow Like Whisky

Tag: AI

  • Claude 4.5 Opus’ Soul Document

    —

    by

    robert.adlington
    in

    Claude 4.5 Opus’ Soul Document — Read on simonwillison.net/2025/Dec/2/claude-soul-document/

  • The space of intelligences is large

    —

    by

    robert.adlington
    in

    LLMs are shaped a lot less by biological evolution and a lot more by commercial evolution. It’s a lot less survival of tribe in the jungle and a lot more solve the problem / get the upvote. LLMs are humanity’s “first contact” with non-animal intelligence. Except it’s muddled and confusing because they are still rooted…

  • Generating an Infographic with Nano Banana Pro aka gemini-3-pro-image-preview

    —

    by

    robert.adlington
    in

    I’m starting a detailed infographic about the Datasette project. I plan to use Google Search to gather specific information about its core processes, which I will then visualize in a multi-panel diagram. — Read on simonwillison.net/2025/Nov/20/nano-banana-pro/

  • Chatbot as Digital Coworker

    —

    by

    robert.adlington
    in

    The era of the chatbot is turning into the era of the digital coworker. To be very clear, Gemini 3 isn’t perfect, and it still needs a manager who can guide and check it. But it suggests that “human in the loop” is evolving from “human who fixes AI mistakes” to “human who directs AI…

  • En abandonnant l’écriture à l’IA, nous risquons de nous empêcher de réfléchir

    —

    by

    robert.adlington
    in

    si écrire, c’est penser, un texte écrit avec l’assistance de ChatGPT ne matérialise-t-il pas les pensées du générateur de texte, plutôt que les nôtres ? Du contenu soi-disant neuf Et à quoi pense donc ChatGPT ? Heureusement, à rien : on l’a dit, le programme se contente de régurgiter les textes figurant dans son corpus d’entraînement, en en…

  • ChatGPT Does Not Have Opinions

    —

    by

    robert.adlington
    in

    ChatGPT is not a human, and, unlike teachers & librarians & scholars, ChatGPT does not have a consistent, legible worldview. In my experience, it readily agrees with any premise you hand it — and brings citations. It may have read every article that can be read, but it has no real opinion — so it is not a credible…

  • Interleaved Thinking

    —

    by

    robert.adlington
    in

    Interleaved thinking is essential for LLM agents: it means alternating between explicit reasoning and tool use, while carrying that reasoning forward between steps.This process significantly enhances planning, self‑correction, and reliability in long workflows. — Read on simonwillison.net/2025/Nov/3/minimax/

  • MCP as Secure Gateway

    —

    by

    robert.adlington
    in

    Instead of a bloated API, an MCP should be a simple, secure gateway that provides a few powerful, high-level tools: download_raw_data(filters…) take_sensitive_gated_action(args…) execute_code_in_environment_with_state(code…) In this model, MCP’s job isn’t to abstract reality for the agent; its job is to manage the auth, networking, and security boundaries and then get out of the way. This makes…

  • The evolution of Albanian AI governance – Marginal REVOLUTION

    —

    by

    robert.adlington
    in

    Rama said the “children,” or assistants, will record everything that happens in parliament and keep legislators informed about discussions or events they miss. “Each one…will serve as an assistant for them who will participate in parliamentary sessions, and will keep a record of everything that happens and will suggest members of parliament. These children will…

  • Setting up a codebase for working with coding agents

    —

    by

    robert.adlington
    in

    For the most part anything that makes a codebase easier for humans to maintain turns out to help agents as well. — Read on simonwillison.net/2025/Oct/25/coding-agent-tips/