Aswin Manohar

To do: 1. Write a screenplay 2. Launch an app 3. Make a film

you must believe you are special and then go so hard, for so long, with such violent refusal to accept any other ending, that reality itself starts running out of ways to tell you no. you must wage a war daily against the ordinary outcome, until the belief you invented out of nothing in a room by yourself has been hammered into the world so many times that it stops being a claim and becomes reality.

Do automated evals actually work? I and @HamelHusain spent last few weeks testing the auto-evals efficacy of different evals platforms. We took 100 real production traces from an apartment-leasing voice agent, manually reviewed the failures, masked the labels, and asked different systems to do the same error-analysis task and discover failure modes. We tested dedicated eval platforms like Braintrust, Arize, and LangSmith, along with ChatGPT, Codex, Claude Code, and Factory Droid. Full post here: parlance-labs.com/blog/posts/aut…

@saaarrth @airindia every other airlines does it with their own cultural roots.

wtf @airindia 😭

Every enterprise will have its own model-harness-sandbox-eval flywheel with token value per watt optimization. This is the future. Simple reason: tacit knowledge about the domain and customers and their workflows that the company uniquely understands and has built trust around.

@mattbotvinick @Anthropic Nobody cares.

We're building a team at @Anthropic focusing on AI and the rule of law. We've made our first hires, and are now opening up a new research engineer role. We're looking for people with advanced technical skills, including AI/deep learning/NLP, full-stack development and data science, paired with training or experience in law, government, political science, or a related field. If this is you, or a friend, please get in touch. job-boards.greenhouse.io/anthropic/jobs…

@thede_plandude Pretty cool, would love to try!

Sometimes when I'm on a call I wanna explain stuff using diagrams. Its too much of a pain to pull up excalidraw and share screen. So, I'm making this browser plugin for myself. Draw, Edit, Move, Scale objects on screen while you talk.

@mattpocockuk I just love this. It feels the right move after a nice agentic coding session. The questionnaire in each lesson was a great idea. Thanks, Matt.

/teach now creates reusable components which get improved/developed as you learn more More token-efficient, and builds momentum Look at these sweet cubes:

Damn, so I can have two subscriptions and one harness?

Here are my 7 phases of AI-powered development. I've been thinking that the pre-PRD phase needs more structure. You need to figure out the shape of the design tree first, before then walking down it with higher-fidelity prototypes. In other words, /grill-with-docs needs to change again IMO

I learnt a long time ago that there is a big difference between making a living and making a life. In the times to come, AI will get increasingly better at the skills that we've used to make a living. Our imperative will be to instead make lives. Not artificial lives. Or artificial lives. But our own lives and of those we love. As machines get better at answering, and solving what they are asked, our work is to get better at asking, at making, creating, and deciding which questions are worth a life, and refusing to outsource that.

@Avinash_R13 I think about it very often. I had this feeling after watching Nobody & Nobody 2. Kamal can do so many great things and all he has to do is work with the new promising directors.

I am so tuned to imagine 'what if it was #KamalHaasan' in every great performance that I watch nowadays... After watching what #NaseeruddinShah did in #MainVaapasAaunga, I can't help but be pretty furious about the last decade of KH being lost to Bigg Boss, politics, and whatever he did after #Vikram Come on... This is what we, or rather I want to see you do. I know you have done soooooooo much already, but you can do soooooooo much more. The could-haves and would-haves are already excruciating. Naseeruddin Shah has given two GOAT performances in the past two months. I so wish you would come back to your true love and passion. Soon. Nandri, Ungal, Naan

There is a lot of noise in the podcasting and influencer space. One thing that always works for me in filtering the voices that I'd like to hear is if they have some proven controversy attached to their name. Surprisingly, they always have one.

Me prompting Claude fable 5 at 3 am be like:

Tokenmaxxing is bad. The real skill is to do what you want with minimum tokens.

I've had random people forwarding me AI research they've been doing. They're not AI researchers. They're just talented hobbyists, usually with a very strong math background, who decided to push on it. Some of the stuff is honestly pretty goddamned impressive. I'm wondering if the time has come for AI research to break out of the labs. It feels like it's just going to emerge spontaneously.

Steps to become a senior programmer: 1. Install my /teach skill npx skills add mattpocock/skills --skill teach 2. Create a new working directory on your laptop mkdir junior-to-senior cd junior-to-senior 3. Kick off your coding agent in the directory claude 4. Copy this prompt /teach me how to be a great strategic programmer. My opinion is that AI is eating 'tactical, on-the-ground' programming. The day-to-day work of a developer involves not only coding, but also planning, QA, codebase design, and much more. I'm interested in learning the strategic skills - that, in a previous era, would take me from junior to senior - but in this era are table stakes. 5. Paste it into the coding agent Below is an example of what the first output will look like. I used Opus 4.8, medium effort. 6. Continue working with the agent until you're a senior

the anthropic co-founder jack clark advice that stuck with me: read the primary material. not the summary. not what the ai said about it. the actual thing. form your own opinion first. then ask the model. never the other way around. keep practices in your life where it’s just you against the world ~ a sport, an instrument, reading, building something with your hands. spaces where the algorithm can’t mediate what you learn about yourself. and don’t defer to AI even when it’s usually right. especially then, actually. that’s precisely when the habit forms. the people who won’t get eaten by this moment are the ones who stayed hard to replace. not because they avoided the tools but because they kept the parts of thinking that make the tools worth using.

The new moat in the agent era is being the tool agents reach for. A coding agent doesn’t reinvent a database. It wires up Supabase. The best devtools companies will make themselves obvious to agents: easy to find, easy to reason about, easy to wire up. Devtools are entering a golden age, but only for companies that realize they’re selling to agents now, not just humans.

Introducing Flows Agent in ElevenCreative. Describe what you want to create and the agent builds your entire workflow - selecting models, creating nodes, wiring connections, and running the generations.

insane amount of alpha in having codex with computer use go through your twitter bookmarks and turn them into a navigable database