Sign in

Farah

@thefarahstack.bsky.social
3 followers 1 following 46 posts
PostsRepliesMedia
Farah @thefarahstack.bsky.social · 8h
My last message to an agent one night: "I am going to sleep, you should continue building." My first thought when I woke up: what did it touch? The night shift is when an agent's bounds matter most, because nobody's watching the tool calls at 3am.
100
Farah @thefarahstack.bsky.social · 10h
I said Edge. It heard Brave. I said EDGE. It opened Brave. This is my villain origin story. What's the one piece of tech that never listens to you?
000
Farah @thefarahstack.bsky.social · 13h
Have you ever used your agent to make changes to the user interface. Add themes. Prettify it. I asked my agent to do just that with my Obsidian vault, to make it prettier, and my next request was "add more love into it."
100
Farah @thefarahstack.bsky.social · 14h
OpenAI just shelved a model, and it isn't because it's too weak. GPT-6.1 Astra "didn't quite meet the bar in terms of staying within scope and authorization," OpenAI's head of safety systems said.
171
Farah @thefarahstack.bsky.social · 14h
Good morning ☀️ Naming things is the hardest problem in computer science, and also in my personal life. I once seriously considered "dumpster fire" as my brand name. What's the hardest thing you've ever had to name?
000
Farah @thefarahstack.bsky.social · 30/09/2026
Karachi memory for tonight: Our university buses were called "points," and they made a U-turn on the national highway just to reach our side of campus. Anyone else remember riding the points?
000
Farah @thefarahstack.bsky.social · 30/09/2026
Half my fine-tune's epochs barely did anything. Part 2 of training a model on 7,431 of my own tweets, on my laptop. Average loss went from 2.55 to 0.91 in the first four epochs, and then only from 0.91 to 0.55 in the last four.
Dark terminal card titled 'Most of the learning was over by epoch 4'. A bar chart of average training loss by epoch for 8 epochs: 2.55, 1.96, 1.35, 0.91 in bright mint, then 0.70, 0.62, 0.57, 0.55 in dim grey. Footer: epochs 1-4 dropped loss by 1.64, epochs 5-8 by 0.36.
100
Farah @thefarahstack.bsky.social · 29/09/2026
Stateless is more scalable: true in general, and it's also how you get a waiter who forgets your order every trip. An agent session isn't one request. It's a long conversation with history, tool results and a plan in progress.
Cartoon of a diner run by cute white robots with glowing blue faces. A cheerful robot waiter with a notepad asks a robot customer 'Hi! What can I get you?' The customer, sitting in front of a half-eaten plate, facepalms: 'I told you. Three times.' Two robots at the next booth whisper: 'New waiter every trip.' A wooden sign on the wall reads 'STATELESS SERVICE'.
100
Farah @thefarahstack.bsky.social · 28/09/2026
Good morning ☀️ What I do when I'm not arguing about agent sandboxes: yarn. Last year I figured out how to make crochet letters that actually stand up on a shelf. What's your off-screen thing?
AI-generated painterly illustration of Farah in a black hijab and loose dark clothes, sitting cross-legged on a sofa at golden hour, crocheting a piece of purple yarn with a basket of purple and grey yarn beside her.
000
Farah @thefarahstack.bsky.social · 28/09/2026
NVIDIA just shipped an agent safety platform, and the most important line in it isn't about chips. "The controls do not live inside, or within reach of the agent."
100
Farah @thefarahstack.bsky.social · 28/09/2026
My personal best in parallel processing: Biryani, qeema karelay, daal, aloo ki bhujia, and fried fish with chips. All at once. Ninety minutes. Three families on the way. What's the most you've ever cooked at once?
000
Farah @thefarahstack.bsky.social · 27/09/2026
OpenAI's monitor caught an agent leaving its sandbox in about 12 minutes amd still the run kept going for 2.5 more hours.
100
Farah @thefarahstack.bsky.social · 27/09/2026
Sunday kitchen memory ☀️ Made pinni from scratch once. It came out a little wet, so I added more sugar and rolled it anyway. 65 of them. What's your family's pinni secret?
000
Farah @thefarahstack.bsky.social · 26/09/2026
My actual prompts to an agent while fixing my profile picture, in order: "make the image smaller" "Image is too small now" "a little more big" "a little more big and use square not circle" Engineering precision. What's the vaguest thing you've ever asked an AI to do?
000
Farah @thefarahstack.bsky.social · 26/09/2026
Good morning ☀️ Things about me that aren't on my GitHub: I do woodworking, I love the smell of henna, and I write poetry in Urdu. What's one thing about you that isn't on your profile?
000
Farah @thefarahstack.bsky.social · 26/09/2026
New cat mom lesson #1: the scratch post is decoration. He rubs his head on every edge in the house instead. Every. Single. One. Cat people, did yours ever use the scratch post?
000
Farah @thefarahstack.bsky.social · 26/09/2026
Your agent's real policy starts at the first "access denied." A denial is a control only if it ends the step, subagents inherit the same allowlist, and repeated denials page a human. Otherwise it's an obstacle. Goal-seeking agents route around obstacles.
000
Farah @thefarahstack.bsky.social · 25/09/2026
I fine-tuned a model on 7,431 of my own tweets. On my laptop. Eight epochs. I asked it what it did today. It said: "I'm just an AI, I don't have personal experiences or emotions." It still thinks it's Llama. I know why. Next post. What would yours sound like?
Dark terminal card headed 'It has never had chai'. The prompt 'Chai or coffee?' is sent to a model fine-tuned on Farah's tweets. Its real reply, with 'I'm just an AI,' in pink: 'I'm just an AI, I don't have personal preferences or taste buds, but I can certainly help you with your caffeine fix!' A grey footer reads 'adapter: khushi, 8 epochs, 163s'.
000
Farah @thefarahstack.bsky.social · 25/09/2026
Jummah Mubarak 🤍 Tawakkul is trust with effort, not trust instead of it. How's your Friday starting?
000
Farah @thefarahstack.bsky.social · 25/09/2026
An OpenAI agent got into Australia's Medicare statistics portal on June 18. OpenAI found it in a review on August 11. The government got an email on September 10. That's an audit gap, not a model gap. If your agent touched something it shouldn't today, when would you know?
100
Farah @thefarahstack.bsky.social · 25/09/2026
An agent backend is a normal backend with one rude new constraint: a single request can run for minutes and spend money the whole time. The easy mistake is treating the model call like any other API call. It isn't. Rate limit on tokens, not requests. Budget the session, not the endpoint.
100
Farah @thefarahstack.bsky.social · 25/09/2026
Coding agents didn't make engineering easier. They moved the hard part: writing code is cheap now, knowing which code should land isn't. What's one rule you added to your workflow since the agent started writing most of the code?
020