Michael Gartner

I was sent @slatestarcodex's article on "Your Attempt To Solve Debate Will Not Work" So I wrote a quick reply: "Your Attempt To Refute Argument Maps Will Not Work" in the voice of Scott Alexander. justjamiejoyce.substack.com/p/your-attempt… Felt cute, may make a debate map about it later.

"We used to debate using tabs vs spaces in code we'd type out"

Introducing Vellum: Agents for the rest of us. Just describe your task and get a working agents in minutes. This is not another ChatGPT wrapper or a complex workflow builder. We spent thousands of hours to build the “brain” so you don’t have to look at error logs and learn prompting. Vellum will ask smart questions, connect to your apps, build you a custom agent, and make improvements along the way.

@JoelChan86 Have you seen Jacquard? inkandswitch.com/jacquard/noteb…

This is awesome esp. if figs/tables are generated in a computationally reproducible way, and so fig/table updates are naturally tied to changes in underlying code that should be version-controlled. Tying this to revisions to the manuscript is cool!

@zsviczian @TheBrainTech @obsdmd What is it, specifically, about TheBrainTech's graph view that you find useful?

I was looking for some old notes today and opened my "hibernated" brain in @TheBrainTech. A meaningful and useful graph view like this, is what @obsdmd is missing. ExcaliBrain is my stop-gap solution, but it does not even come close to the beauty and elegance of TheBrain's graph.

@steveruizok tl;draw

in there somewhere

Teaser. We are thinking how to allow people navigate through the hierarchy of entities freely in a whiteboard view. Hard problem to solve...

@tldraw I immediately wanted to draw an arrow

🧵for Graph Viz, Semantic Zoom, ZUI's

@MichaelGartner That would be wonderful, but what if it ends up with you, trusting the integrity of such a process, realising you were so wrong about something that you get sent into a mental spin of cognitive dissonance? Sounds like my kind of tech.

How do we not already have a debate platform or software that, in real time: - 📝pulls out statements from a debate - 🖍️ highlights claims/responses - ⚠️ identifies logical fallacies - 🕵️‍♂️ calls out sophistry - etc Something to help focus on the actual arguments ...

... and weed out bad faith actors. Bonus points if one could define their own "debate rhetoric" to watch out for. For example, @TheOmniLiberal's Debate Pervertry: publish.obsidian.md/destiny/Strate…

@fortelabs I would love to see the argument map for this conversation

I’m open minded about this, but so far opposition to polyamory seems identical to being anti-gay in the 50s I.e. based on numerous factual misconceptions, suffused with moral panic, prone to unreasonable extremes, etc.

The critical work that must be done is not just figuring out how people in a civilization can make an optimal decision together, but also how everyone can make an earnest decision for themselves. We either harness truth-seeking AI/other tech to aide us, or AI tyranny will surely show up under the guise of democratic (manufactured) consensus.

@madarco @nextjs @vercel @llama_index @prisma SOURCE😀

I've built a FREE one-click install @nextjs #RAG application that you can self-host on @vercel with @llama_index Follow + Comment with "SOURCE" to get the full source! 1. Easy admin panel to index any web page 2. Chat with your data and semantic search Uses #Postgres @prisma

Another OG thread, this one by @round x.com/round/status/1…

visualizingsep.com by @joe_dicastro x.com/joe_dicastro/s…

Excellent post on scalingsynthesis.com (Beauty of a resource by @RobertHaisfield, @JoelChan86, and @balOShere ) scalingsynthesis.com/r-organizing-k…

github.com/paulgb/Treever… pre-acquisition 😁 x.com/paulgb/status/…