Brent

This account is verified. Learn more

@bigfarmer666

Operations management. Agentic systems, BJJ, degen powerlifter.

In the trenches working it out.

Bit of a personal diary.
Follow along. Or don’t.

Brunswick Melbourne via Tassie Joined September 2023

483 Following 122 Followers 2.7K Posts

Latest posts 5

Brent @bigfarmer666 Oct 17

Jibber-jabbering to my @bot CoS to pull my @NotionHQ task list and queue jobs for my knowledgebase ingest @bot, that then calls @herdrdev via @bot shepherd to run today’s ingests - all while monitoring the CLIs with @theSSHHIP on my phone and sinking beers at the pub.

Absolutely ridiculous (and perfect) workflow.

Brent @bigfarmer666 Oct 15

Now with new fancy name and more official accounts

x.ai/bot/cogaww1OUG…

x.ai sXsi Radar by Brent Watches the Grok Bot and Cursor team accounts on X and sends you one morning digest of what's worth knowing, grouped by product, with links, at a time...

Brent @bigfarmer666 Sep 19

Yep. I’m fighting 30 years worth of key-person risk at my org.

Had to go very slow, but in tiny increments I’m extracting knowledge out of people’s heads.

My first move was to solve the ridiculous state of the comms.

Mark Ajzenstadt Verified account @mardehaym Sep 18

x.com/i/article/2101…
View quoted post on X

Brent @bigfarmer666 Sep 19

Parallel search with @typesafeai Jev reranking/disambiguation is go.

A bit fiddly but we got there.

System of record/ERP MCP @bot 1

Knowledge base - QMD hybrid/vector search @bot 2

Gmail @bot 3

Next up: front end classifier/router for Chief of Staff @bot search calls

Brent @bigfarmer666 Sep 18

This typesafe.ai is mental.

Never thought I’d get so excited over a classifier

typesafe.ai Home - TypeSafe AI TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access.

Diogo Almeida Verified account @CompleteSkeptic Sep 15

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?

I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev

• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions

AFAICT the shortest path to AI-based economic revolution

View quoted post on X