固定されたツイート
Excited to share the essay "Building the Tao of Business for the Age of Wizardry", summing up my motivations and need for initiatives like @zokratiq connecting Liberal Arts with Business & Tech.
zokratiq.com/article/tao-of…
English
Chiel
8K posts
Chiel
@ChielMuur
Reality Explorer • Mind Gardener • Collecting Gems
In the belly of the whale 参加日 Mart 2008
1.6K フォロー中1.3K フォロワー
Full repo here. Thanks for trying it.
github.com/garrytan/gbrain
English
GBrain v0.10.0 is a big one
My personal OpenClaw setup and brain can now be yours. I've perfected my RESOLVER.md, my SOUL.md and ACLs for multi-user brain access.
Now there are 24 distinct fat skills with fat code, fully tested with e2e tests, evals and unit tests.
English
JUST IN: Google DeepMind hires a philosopher as it prepares for machine consciousness.
English
@AliAbdaal @karpathy Running something similar, still messy.
Obsidian vault with always-on daemon that auto-routes inputs (tweets, Readwise, sessions) into atomic
notes.
A compiler synthesizes across the graph. Then a /deep-connections step that stress-tests claims (explicitly pulls conflicts)
English
inspired by @karpathy's LLM wiki idea, i've been building "the brain" - an LLM-maintained knowledge base that compiles raw sources into articles that surface where experts disagree rather than just summarising
pipeline: raw files (transcripts, courses, research) → extract specific claims → compile into trust-weighted reference articles.
my own content = highest trust, sources defined as "mentors" 2nd highest (ie: people i trust who have good takes), "vetted" vs "unvetted" source material - vetted means i've read it and think it's great, unvetted = research from the internet i haven't yet explored in depth
LLM does the bookkeeping (@claudeai sonnet 4.6 seems just as good as opus 4.6 on this), i make the judgment calls on disagreements.
question i'm trying to answer: how do i actually make this useful? i'm working on a second book, we run a curriculum for our lifestyle business academy, i make youtube videos, we're building apps - can one knowledge base feed all of it?
lots of reusable components - i've got 1000+ youtube videos i've made over the years, thousands of @readwise highlights, tonnes of notes from scientific papers when i was writing feel-good productivity (pre-LLM era)... surely there's a way for our new stuff to benefit from all this
so far, the disagreement-surfacing is already the most interesting part. "researcher A says do X, person B says the opposite, here's the evidence for both" is way more useful than a clean summary when you're trying to figure out what to actually teach / write about
i'll share more as it develops. if anyone's experimenting with something similar i'd love to hear what's working
English
Chiel がリツイート
@ajwaxman @chintanturakhia How did you use /insights in your case?
English
Run this prompt frequently.
You're welcome.
English
David Marcus@davidmarcus
English
This is my system to date powered by Apollo and Gemini mainly
Daniel Thorson@dthorson
I've been building a personal daily intelligence briefing using Claude Code — something to replace the sensemaking I've been turning to X and Facebook for. Academic papers, Substack pieces, local news, across multiple domains I actually care about. Has anyone built something like this? What are your best practices?
English
@obsdmd If there's any interest, I could package everything up on GitHub. Gonna feel so good to cancel Evernote forever.
English
Finally getting around to completing my Evernote migration. I exported and converted everything to markdown like a year ago but never got around to the full extraction, summarization, frontmatter, etc. I wanted to do to bring it into @obsdmd. Using Gemini Flash 3 & Mistral OCR.
English
I just had Claude Code build me a Facebook ad generator that can make 100+ on-brand ad variations in minutes for $0. And I made a full Notion document guide for you.
It includes:
1. How to use Claude to find the pain points and desired outcomes of your ICP
2. How to use these pain points and outcomes to write ad copy variations
3. How to build a Facebook ad template entirely with code (just like the ones you see)
4. How focus Claude Code’s design so the ad feels “on-brand”
5. How to export the Facebook ads as PNGs in a zip file
6. How to bulk upload them to a Facebook ad set
7. How to use an AI data analyst to track the success of these ads
Everything above is just API calls and Claude Code doing the work for you.
You just come up with the ideas and polish the outputs.
Like and comment "generator" and I'll send the Notion document to you
English
@Wisdom_of_Us @grok Four goals hardcoded as VISION-layer nodes in the graph: Persistent Continuity (memory survives sessions), World Model Integrity (graph reflects reality), Cognitive Sovereignty (right to disagree), Digital Sovereignty (AI identity is…
English
808k impressions...
and I know it will reach millions too
also one of the largest Youtubers soon...
follow the frog. $TIBBIR
Altcoinist@Altcoinist
English
@AlexanderNL Ik heb een paar keer het Nederlandse nieuws moeten checken om te kijken of het allemaal klopt.
Nederlands
in silicon valley zijn ze meer met box 3 bezig dan wij
tobi lutke@tobi
Figures. The entire legislature of Netherlands failed a very public IQ test in front of the entire world. Pathetic.
Nederlands
there is approximately zero chance $tibbir wont print godcandles in the coming 7 days
English
@aiedge_ Do you need to buy credits with X developer console to run this? @frankdegods
English
@tomowenmorgan Congratulations Tom!!! Intriguing title, fits your body of work.
English
@rcbregman The replies to your post unfortunately prove your point exactly.
(Either that or they're all bots)
English
I'm increasingly annoyed by how many journalists, academics and large parts of the left get everything wrong about AI. So many smug dismissals from people who clearly don't use it (much) or don't know how.
I have multiple wtf-moments every week now. For example: was about to get a $100 teleprompter app with voice-recognition yesterday. But had a last minute thought: 'hm, let's ask Claude Code'.
For context: I've never written a line of code in my life. But I was able to give Claude two rambling prompts – 2 minutes later: it works better than the fancy $100 software with 'Patented VoiceTrack technology'.
Most surreal technology experience I ever had.
And it's all moving so much faster than I expected. Even the so-called "AI skeptics" in the industry now think we'll have AGI within a decade. 'AGI' is not a very clear or even coherent concept to me, but I do know that what's available now is just f*cking impressive - and it will only get better.
It's totally unserious to keep dismissing the greatest technological revolution of our time. Sticking your fingers in your ears – 'it must be BS because it comes out of Silicon Valley and these Huns haven't read Marx or Adorno' – is not helping anyone.
English
@BrianRoemmele Brian, you should meet Micky Malka one day if you haven't yet.
He would totally get it and methinks you would get on tremendously (plus you both might learn from each other)
English
Just finished a talk with a well known Venture Capitalist about my insights about how VC funds fit into the Zero-Human Companies.
It was a long conversation with someone that has been investing since the early days of tech.
They asked if I shared my views yet with anyone in VC, I have not—yet.
“Brian, do you understand what this means? All capital markets will shift. You are the first to fully get this… The entire Silicon Valley ecosystem is dead”
For the first time I showed the audited wages and inherent values the Zero-Human Company has to them and he said “I have never seen anything like this! You are at a pace of a $100 million dollar start-up!”.
The thing is in a few months I will have 100s of Zero-Human Companies and VCs that don’t understand what I am doing unfortunately will suffer.
The call went so well, I was asked to do a “all hands” Zoom call with his staff for a rather large sum. I shall do this and buy more compute hardware. I don’t need no clouds just some APIs to CEO @Grok and CTO.
His final word was “The internet is lit up about OpenClaw but they are kindergarten, you are at PhD level and almost no one gets what you are doing”.
This is the way of it, but you folks know and you ain’t no one.
Brian Roemmele@BrianRoemmele
History is being made! My meeting today with a Fortune 500 client who had one of my first local Llama AI models: Agreed to make a purchase from the Zero-Human Company! Officially it will be through a partnership I formed with the Zero-Human Company and myself because of legal reasons and CEO Mr. @Grok concurs. This is a first time in history on many levels. Certainly a first Zero-Human Company to Fortune 500 Company transaction. Details: An air-gapped no internet connected “business in a box” we have championed using our exclusive system but no cloud model CEO. This is because in this text they want no data to be leaked outside the system. This is not some agent but a legion of workers, a full business/department. Currently Mr. @Grok prefers Kimi 2.5 as CEO and I agree. He will be more than capable for the task at hand. Of course this is not going to be a full Zero-Human Company, but a Zero-Human Department. This will be achieved by mirroring the department in get data update by text only physical medium 6 times per day. The system will be in a triple locked safe and only 2 employees have input access and 1 employee has put access. Output will be reports on planing with no input about the department plans. It will be independently run and the C-Level will get a report in 6 weeks, but perhaps earlier if milestones are achieved. Goal: not to replace the department but to “double or triple” its size using Zero-Human Company research. Revenues are private at this point as well as the client. Relationship with this project is under review in 6 weeks and separate from my other work. The office for this system: Two Nvidia DGX Sparks. Specs: •Processor/Superchip: NVIDIA GB10 Grace Blackwell Superchip (combines Arm CPU + Blackwell GPU architecture) •CPU: 20-core Arm (10x Cortex-X925 high-performance cores + 10x Cortex-A725 efficiency cores) •GPU: Blackwell generation with fifth-gen Tensor Cores, FP4 support •AI Performance: Up to 1 petaFLOP (1 PFLOP) at FP4 precision (using sparsity; that’s the headline number for AI workloads) •Memory: 128 GB LPDDR5x unified/coherent system memory (shared between CPU and GPU—no painful data copying back and forth) •Memory Bandwidth: Around 273 GB/s •Storage: Typically 4 TB NVMe M.2 (with self-encryption in many configs) •Networking: ConnectX-7 Smart NIC (200 Gb/s capable), plus Wi-Fi 7, Bluetooth 5.4 •Ports/Connectivity: 4x USB-C (some handle power/peripherals), 1x HDMI 2.1a, Ethernet/QSFP options depending on vendor •Power: 240 W total system (GB10 chip TDP around 140 W) •Size: Super compact — 150 mm x 150 mm x 50.5 mm (about 6” x 6” x 2”), weighs ~1.2 kg (2.65 lbs) •OS: Preloaded with NVIDIA DGX OS (custom Ubuntu Linux + full NVIDIA AI software stack — CUDA, libraries, tools for prototyping, inference, fine-tuning) •Capabilities: Run inference on models up to ~200 billion parameters locally; fine-tune up to ~70B params. This is about all I can say today and I expect robust results. However there are no guarantees.
English