# Confidently Incorrect — Artificial Insights | Envisioning Source: https://www.envisioning.com/newsletter/077-confidently-incorrect --- ![](/_next/image?url=https%3A%2F%2Fres.cloudinary.com%2Fenvisioning%2Fimage%2Fupload%2Fv1771331166%2Fnewsletter%2Fissue%2F077-confidently-incorrect%2Fcover%2Fcover-a4fc75c6.jpg&w=3840&q=75) [Back to newsletter](/newsletter) # Confidently Incorrect Issue 077 · November 4, 2024 0 Happy Monday and welcome to your weekly download of AI opinions, news and memes. I am short on personal commentary this week but think you might enjoy these new-to-me ways of using AI throughout the day: **Ask quick questions from my computer** using [Raycast](https://www.raycast.com). The desktop ChatGPT app also features a keyboard triggered chatbar, but I find Raycast fast and seamless for most simple queries. ![](https://res.cloudinary.com/envisioning/image/upload/v1771331696/newsletter/issue/077-confidently-incorrect/inline/img-001-122ff442.gif) Ask Claude or ChatGPT to **create scripts** for every kind of data operation I can think of, like merging or organzing structured data files. Maybe you don’t spend as much time as I do manipulating data, but the application is general. Scripts, particularly in Python, are a powerful way of managing information. ![](https://res.cloudinary.com/envisioning/image/upload/v1771331697/newsletter/issue/077-confidently-incorrect/inline/img-002-ae25bb3b.gif) Using ChatGPT’s advanced voice mode to help me **practice Dutch** (or almost any language). Duolingo only goes so far, so instead of bothering locals with my poor Dutch, I can spend a couple of minutes daily with an infinitely patient language coach who both corrects me and gives me some confidence to actually speak to strangers. ![](https://res.cloudinary.com/envisioning/image/upload/v1771331698/newsletter/issue/077-confidently-incorrect/inline/img-003-13d0e802.gif) Asking **Claude** to generate interactive explainers from our [AI Vocab](https://www.envisioning.io/vocab/) to help more people understand the underlying concepts. Claude is pretty good at creating web components, especially through dialogue, which can be used to illustrate many, maybe all Vocab terms. ![](https://res.cloudinary.com/envisioning/image/upload/v1771331699/newsletter/issue/077-confidently-incorrect/inline/img-004-6ec00a0d.png) I hope this triggers insights into how to use similar tools yourself! Until next week, MZ * * * ![](https://res.cloudinary.com/envisioning/image/upload/v1771331700/newsletter/issue/077-confidently-incorrect/inline/img-005-8197bf6a.heic) * * * #### Unintended Consequences of AI's Evolution (20 min) Technical talk about the implications of generative UX, which touches on some of my favorite topics: Centaurs and unintended consequences of technology. ![YouTube video thumbnail](https://img.youtube.com/vi/apZXpgU_klw/hqdefault.jpg)Play on YouTube [Watch on YouTube](https://www.youtube.com/watch?v=apZXpgU_klw) > ._..good Sci-Fi predicts cars, great sci-fi predicts traffic._ * * * #### GitHub Spark Showcase (1 min) GitHub stepping up the game. Good time to be a coder. * * * #### Physical AI (3 min) Hype-ish but technically accurate video by NVidia about robotics, which they call physical AI, and how they are enabling it. ![YouTube video thumbnail](https://img.youtube.com/vi/uhLDHA9skFk/hqdefault.jpg)Play on YouTube [Watch on YouTube](https://www.youtube.com/watch?v=uhLDHA9skFk) * * * #### Two Possible Futures (30 min) If, like myself, you have been meaning to read [Dario Amodei](https://darioamodei.com/machines-of-loving-grace) and [Sam Altman's](https://ia.samaltman.com) recent essays but haven't found the time yet, here is an excellent (human) podcast discussing their takes and differences (the first 30 min or so). Here is a NotebookLM version of the same comparison of both essays with minimal instructions. * * * #### Planning in AI (1h) How we reach generality: OpenAI research scientist Noam Brown on planning & search in AI. Dense and technical and fascinating. ![YouTube video thumbnail](https://img.youtube.com/vi/eaAonE58sLU/hqdefault.jpg)Play on YouTube [Watch on YouTube](https://www.youtube.com/watch?v=eaAonE58sLU) > _Just by adding search, we were able to do the equivalent of scaling the models by 100,000×_ * * * #### The Leap to 10 Trillion Parameters (30 min) Feed the hype. Great conversation from YC. ![YouTube video thumbnail](https://img.youtube.com/vi/lbJilIQhHko/hqdefault.jpg)Play on YouTube [Watch on YouTube](https://www.youtube.com/watch?v=lbJilIQhHko) > _...if we make the models smart enough that they can actually do original thinking and deep analysis with correct logic... maybe we'll finally invent it all._ * * * #### Envisioning Vocab Don’t miss the generative knowledge graph of [our AI terms](https://www.envisioning.io/vocab/). Very experimental but fully functional. A navigable graph of all (800+) vocab terms where connections are made based on the similarity between nodes, creating a sort of map of all AI related concepts out there. Don’t miss it: [envisioning.io/vocab/map/](https://www.envisioning.io/vocab/map/) ![](https://res.cloudinary.com/envisioning/image/upload/v1771331702/newsletter/issue/077-confidently-incorrect/inline/img-006-bf97bff8.png) ![](https://res.cloudinary.com/envisioning/image/upload/v1771331703/newsletter/issue/077-confidently-incorrect/inline/img-007-ac0635f5.png) [envisioning.io/vocab/map/](https://www.envisioning.io/vocab/map/) We are also experimenting with a timeline view in the shape of a sunflower. I asked GPT to determine the year of invention for each of the terms, and have used this data to order the entries by age. Check it out: [envisioning.io/vocab/timeline/](https://www.envisioning.io/vocab/timeline/) ![](https://res.cloudinary.com/envisioning/image/upload/v1771331704/newsletter/issue/077-confidently-incorrect/inline/img-008-3643a814.png) [envisioning.io/vocab/timeline/](https://www.envisioning.io/vocab/timeline/) We are thinking of applying this methodology to other technical fields. LMK if you have suggestions. _You are receiving this newsletter because you signed up on [envisioning.io](https://envisioning.io) or Substack._ * * * ## More from Artificial Insights --- ![Nov 11, 2024 · Issue 078](https://res.cloudinary.com/envisioning/image/upload/f_auto,q_auto:best/v1771331167/newsletter/issue/078-everything-is-obvious-in-retrospect/cover/cover-82233f57.jpg) Nov 11, 2024 · Issue 078 Everything is Obvious in Retrospect [/newsletter/078-everything-is-obvious-in-retrospect](/newsletter/078-everything-is-obvious-in-retrospect) 0 ![Oct 28, 2024 · Issue 076](https://res.cloudinary.com/envisioning/image/upload/f_auto,q_auto:best/v1771331165/newsletter/issue/076-systems-using-systems/cover/cover-a028be24.jpg) Oct 28, 2024 · Issue 076 Systems using Systems [/newsletter/076-systems-using-systems](/newsletter/076-systems-using-systems) 0 --- [View all issues](/newsletter) * * * ## Comments Post commentComments are reviewed before appearing. ## Newsletter Follow us for weekly foresight in your inbox. Sign up