Muhammed Jasim

Building enterprise systems and products that ship. Building @FlowtrailAI. Sharing what I learn.

[Muhammed Jasim @mjasimea]

Playwright MCP or Chrome DevTools MCP?

Different tools, different jobs. Here's how I pick.

𝗣𝗹𝗮𝘆𝘄𝗿𝗶𝗴𝗵𝘁 𝗠𝗖𝗣 — When I need reliable automation. Click this, fill that, take screenshot, move on. It's battle-tested. Predictable. Great for E2E flows where you just need the agent to do the steps.

𝗖𝗵𝗿𝗼𝗺𝗲 𝗗𝗲𝘃𝗧𝗼𝗼𝗹𝘀 𝗠𝗖𝗣 — When I need to understand what's happening. Performance traces. Console errors. Network requests. The agent isn't just clicking — it's debugging.

Simple rule I follow:
If the task is "do something in the browser" → Playwright.
If the task is "figure out why something broke" → DevTools.

#AI #WebDev #MCP

[Muhammed Jasim @mjasimea]

Not everything needs to be an agent.

If you know the steps → write the code.
If the user knows what they want → write the code.
If the logic is predefined → write the code.

Sometimes the best AI system is one where AI does less.

#AI #AIAgents #SoftwareArchitecture #Programming

[Muhammed Jasim @mjasimea]

@AGrillz @ycombinator @riveterhq F25

[Muhammed Jasim @mjasimea]

@unwind_ai_ Agents that actually get better over time — love this concept.

Great work on this.

[Muhammed Jasim @mjasimea]

@AnthropicAI @bunjavascript Amazing. Two of my favorite tools joining forces.

Congrats @jarredsumner - excited to see the great tools coming from the @bunjavascript at @AnthropicAI

[Muhammed Jasim @mjasimea]

Your users have GPUs. Why are you paying for inference?

Browser-based ML is real now. 1.7B models running at 160 tokens/sec. In a browser tab.

No API. No server. No data leaving the device.

What you can build:
→ Real-time transcription
→ Background removal
→ Document summarization
→ Live translation
→ Voice assistants that work offline

The stack: Transformers.js, Chrome Built-in AI, WebLLM, LiteRT.js, ONNX Runtime, TensorFlow.js

Local AI isn't a compromise. It's the better architecture.

[Muhammed Jasim @mjasimea]

Building agents with 50+ tools?

You’re probably burning 100K+ tokens on tool definitions before Claude even reads your message

@AnthropicAI just dropped 3 features that fix this:

⚡ defer_loading: true — tools discovered on-demand
⚡ allowed_callers: ["code_execution"] — batch tool calls in code
⚡ input_examples — teach usage patterns, not just schemas

85% less context. Way better accuracy.

Time to ship some serious agents 🚀

#AI #AIAgents #Claude #Anthropic #DevTools #LLM #MCP

[Muhammed Jasim @mjasimea]

@vercel entering the workflow-automation space is a big move.

The new Workflow Builder turns serverless logic into a visual flow — CRONs, webhooks, queues, AI steps, everything stitched together without extra infra.

https://t.co/YFPbklZVnc

[Claude @claudeai]

Introducing Claude Code on the web.

You can now delegate coding tasks to Claude without opening your terminal.

[Andrej Karpathy @karpathy]

nanochat d32, i.e. the depth 32 version that I specced for $1000, up from $100 has finished training after ~33 hours, and looks good. All the metrics go up quite a bit across pretraining, SFT and RL. CORE score of 0.31 is now well above GPT-2 at ~0.26. GSM8K went ~8% -> ~20%, etc. So that's encouraging.

The model is pretty fun to talk to, but judging from some early interactions I think people have a little bit too much expectation for these micro models. There is a reason that frontier LLM labs raise billions to train their models. nanochat models cost $100 - $1000 to train from scratch. The $100 nanochat is 1/1000th the size of GPT-3 in parameters, which came out 5 years ago. So I urge some perspective. Talking to micro models you have to imagine you're talking to a kindergarten child. They say cute things, wrong things, they are a bit confused, a bit naive, sometimes a little non-sensical, they hallucinate a ton (but it's amusing), etc.

Full detail/report on this run is here:
https://t.co/nWbfKOZLIg
And I pushed the new script run1000 sh to the nanochat repo if anyone would like to reproduce. Totally understand if you'd like to spend $1000 on something else :D

If you like, I am currently hosting the model so you can talk to it on a webchat as you'd talk to ChatGPT. I'm not going to post the URL here because I'm afraid it will get crushed. You'll have to look for it if you care enough. I'm also attaching a few funny conversations I had with the model earlier into the image, just to give a sense.

Next up, I am going to do one pass of tuning and optimizing the training throughput, then maybe return back to scaling and maybe training the next tier of a bigger model.

[Muhammed Jasim @mjasimea]

Adopting @pontusab's ai-sdk-tools at @flowtrailai Clean integration, solid DX. Exactly what we needed.

#AI #DeveloperTools #AIEngineering

[Andrej Karpathy @karpathy]

My pleasure to come on Dwarkesh last week, I thought the questions and conversation were really good.

I re-watched the pod just now too. First of all, yes I know, and I'm sorry that I speak so fast :). It's to my detriment because sometimes my speaking thread out-executes my thinking thread, so I think I botched a few explanations due to that, and sometimes I was also nervous that I'm going too much on a tangent or too deep into something relatively spurious. Anyway, a few notes/pointers:

AGI timelines. My comments on AGI timelines looks to be the most trending part of the early response. This is the "decade of agents" is a reference to this earlier tweet https://t.co/NiSn6jftqq Basically my AI timelines are about 5-10X pessimistic w.r.t. what you'll find in your neighborhood SF AI house party or on your twitter timeline, but still quite optimistic w.r.t. a rising tide of AI deniers and skeptics. The apparent conflict is not: imo we simultaneously 1) saw a huge amount of progress in recent years with LLMs while 2) there is still a lot of work remaining (grunt work, integration work, sensors and actuators to the physical world, societal work, safety and security work (jailbreaks, poisoning, etc.)) and also research to get done before we have an entity that you'd prefer to hire over a person for an arbitrary job in the world. I think that overall, 10 years should otherwise be a very bullish timeline for AGI, it's only in contrast to present hype that it doesn't feel that way.

Animals vs Ghosts. My earlier writeup on Sutton's podcast https://t.co/rSp1noyGBr . I am suspicious that there is a single simple algorithm you can let loose on the world and it learns everything from scratch. If someone builds such a thing, I will be wrong and it will be the most incredible breakthrough in AI. In my mind, animals are not an example of this at all - they are prepackaged with a ton of intelligence by evolution and the learning they do is quite minimal overall (example: Zebra at birth). Putting our engineering hats on, we're not going to redo evolution. But with LLMs we have stumbled by an alternative approach to "prepackage" a ton of intelligence in a neural network - not by evolution, but by predicting the next token over the internet. This approach leads to a different kind of entity in the intelligence space. Distinct from animals, more like ghosts or spirits. But we can (and should) make them more animal like over time and in some ways that's what a lot of frontier work is about.

On RL. I've critiqued RL a few times already, e.g. https://t.co/mYrMFVdVDW . First, you're "sucking supervision through a straw", so I think the signal/flop is very bad. RL is also very noisy because a completion might have lots of errors that might get encourages (if you happen to stumble to the right answer), and conversely brilliant insight tokens that might get discouraged (if you happen to screw up later). Process supervision and LLM judges have issues too. I think we'll see alternative learning paradigms. I am long "agentic interaction" but short "reinforcement learning" https://t.co/2L7FiaoKsw. I've seen a number of papers pop up recently that are imo barking up the right tree along the lines of what I called "system prompt learning" https://t.co/df5mJDdN3C , but I think there is also a gap between ideas on arxiv and actual, at scale implementation at an LLM frontier lab that works in a general way. I am overall quite optimistic that we'll see good progress on this dimension of remaining work quite soon, and e.g. I'd even say ChatGPT memory and so on are primordial deployed examples of new learning paradigms.

Cognitive core. My earlier post on "cognitive core": https://t.co/q2s1ihGy0T , the idea of stripping down LLMs, of making it harder for them to memorize, or actively stripping away their memory, to make them better at generalization. Otherwise they lean too hard on what they've memorized. Humans can't memorize so easily, which now looks more like a feature than a bug by contrast. Maybe the inability to memorize is a kind of regularization. Also my post from a while back on how the trend in model size is "backwards" and why "the models have to first get larger before they can get smaller" https://t.co/6k0FZRGXsb

Time travel to Yann LeCun 1989. This is the post that I did a very hasty/bad job of describing on the pod: https://t.co/fQgqaXPyp6 . Basically - how much could you improve Yann LeCun's results with the knowledge of 33 years of algorithmic progress? How constrained were the results by each of algorithms, data, and compute? Case study there of.

nanochat. My end-to-end implementation of the ChatGPT training/inference pipeline (the bare essentials) https://t.co/SIetgyoKWN

On LLM agents. My critique of the industry is more in overshooting the tooling w.r.t. present capability. I live in what I view as an intermediate world where I want to collaborate with LLMs and where our pros/cons are matched up. The industry lives in a future where fully autonomous entities collaborate in parallel to write all the code and humans are useless. For example, I don't want an Agent that goes off for 20 minutes and comes back with 1,000 lines of code. I certainly don't feel ready to supervise a team of 10 of them. I'd like to go in chunks that I can keep in my head, where an LLM explains the code that it is writing. I'd like it to prove to me that what it did is correct, I want it to pull the API docs and show me that it used things correctly. I want it to make fewer assumptions and ask/collaborate with me when not sure about something. I want to learn along the way and become better as a programmer, not just get served mountains of code that I'm told works. I just think the tools should be more realistic w.r.t. their capability and how they fit into the industry today, and I fear that if this isn't done well we might end up with mountains of slop accumulating across software, and an increase in vulnerabilities, security breaches and etc. https://t.co/8556ESSpyY

Job automation. How the radiologists are doing great https://t.co/FVUI872dkD and what jobs are more susceptible to automation and why.

Physics. Children should learn physics in early education not because they go on to do physics, but because it is the subject that best boots up a brain. Physicists are the intellectual embryonic stem cell https://t.co/p72Elk8lPV I have a longer post that has been half-written in my drafts for ~year, which I hope to finish soon.

Thanks again Dwarkesh for having me over!

[Maximiliano Firtman @firt]

🔴Google introduces LiteRT.js: a new library to run AI models client-side. #WebAI

It shares model format with Android including a PyTorch converter.

[AI SDK @aisdk]

AI SDK 5

Introducing type-safe chat, agentic loop controls, data parts, speech generation and transcription, Zod 4 support, global provider, and raw request access.

[Muhammed Jasim @mjasimea]

🎨 Tried @SuperDesignDev recently — an open-source design agent that runs inside IDEs like VS Code, @cursor_ai, and @windsurf_ai.

It lets you generate UI components, wireframes, and layout variants using prompts, all saved locally in .superdesign/.

No need to switch to Figma or other design tools. The workflow stays within the editor, which is helpful if you prefer iterating in code.

Not something for every use case, but interesting for quick mockups or early-stage UI drafts.

#SuperDesign #OpenSource #AItools #UIDesign #DeveloperTools #AIinDesign #VibeCoding

[Muhammed Jasim @mjasimea]

We've integrated Claude Code into our dev team at @flowtrailai 🚀

Our GitHub Actions integration:
✅ Provides automated, context-aware PR reviews
✅ Manages smart issue assignment and triage
✅ Assigns tasks directly by creating and assigning issues
✅ Writes production code for users

AI building AI, practically 🤖

The future of development feels normal—just another teammate who never sleeps 😄

#ClaudeCode #AIdev #BuildingAI #GitHubActions #SoftwareDevelopment #Automation

[Muhammed Jasim @mjasimea]

Deep Research Agent in ~100 lines using @vercel AI SDK

Powered by 🧠 @OpenAI + 🌍 Tavily + Brave Search
It handles:
→ Smart query generation
→ Multi-source search
→ Credibility checks
→ Report generation

@aisdk #OpenAI #VercelAI #AIagents #DevTools

[Todd Motto @toddmotto]

Installing Prettier for @code was the best thing I've done in a while! Auto-formatting ftw.

[Muhammed Jasim @mjasimea]

@victorhazbun Expecting to be published on this April

[James Ide @JI]

ViewPagerAndroid and ES6 "import" are here in #ReactNative 0.13.0-rc with @martinkonicek's help 🎉🎉🎉
Changelog here: https://t.co/ciZpg52c0z