nummanthinks's profile picture. Applied AI & Agentic Coding
    
CTO at UK FinTech https://www.retailbook.com

Numman Ali

@nummanthinks

Applied AI & Agentic Coding CTO at UK FinTech https://www.retailbook.com

Matt doesn't normally post like this so he means business ✊ Seriously, don't become dependent on one AI Coding Tool Become familiar with the patterns, habits and best practices A new paradigm will come around every few months so make sure you're ready for it

I don't give a shit about what AI coding tool I'm using, and neither should you I care about learning habits that will mean I can pick up any future tool and just GO



Become a Cursor Pro in 12 minutes (I do all of these!)

Learn how to use Cursor Agent with 10 tips: 1. Plan mode 2. Context menu 3. Custom commands 4. Images 5. Duplicating chats 6. Context visibility 7. Usage visibility 8. Keyboard shortcuts 9. New chats 10. Checkpoints And an extra bonus round!



If you're a web developer and want to create advanced ai agent systems/workflows programmatically The new ai-sdk-tools/agent packages from @middayai is absolutely where you should start We've lacked a simple, lowly opinionated agent framework for TypeScript - this is it…

Introducing `ai-sdk-tools/agents`. Multi-agent orchestration for AI SDK v5. Specialized agents, automatic routing, seamless handoffs. 🧵 Thread below with what makes it special ↓



Why isn't anyone talking about the @ChatGPTapp Apps SDK? Thought there would be more hype, more demos It's really cool, I'm trying out all the features and generally really intuitive I believe improvements are needed on the ergonomics but this is potentially a big a shift…


This looks dangerous You must wonder how they did it

🚀 KAT-Dev-72B-Exp becomes the top-1 open-source model to achieve 74.6% on SWE-Bench Verified ⚡ (evaluated strictly with the SWE-agent scaffold). ✨ Try our strongest model KAT Coder directly on StreamLake → streamlake.ai/product/kat-co… ⏳ FREE ACCESS every day! Don’t miss out.…

KwaiAICoder's tweet image. 🚀 KAT-Dev-72B-Exp becomes the top-1 open-source model to achieve 74.6% on SWE-Bench Verified ⚡ (evaluated strictly with the SWE-agent scaffold).

✨ Try our strongest model KAT Coder directly on StreamLake → streamlake.ai/product/kat-co… 
⏳ FREE ACCESS every day! Don’t miss out.…


Building a ChatGPT Apps SDK Playground Repo TypeScript, All Examples, Clone and Build Your Own What would you like to see in it?


Probably the easiest way to get an app on @ChatGPTapp You can even use @nextjs for the full setup


Math is the backbone of everything I expect Gemini 3 to bring quite the game to the coding arena I love the competition between frontier labs It makes innovation come fast and lets us eat new flavours of cake each time 😋

We evaluated Gemini 2.5 Deep Think on FrontierMath. There is no API, so we ran it manually. The results: a new record! We also conducted a more holistic evaluation of its math capabilities. 🧵

EpochAIResearch's tweet image. We evaluated Gemini 2.5 Deep Think on FrontierMath. There is no API, so we ran it manually. The results: a new record!

We also conducted a more holistic evaluation of its math capabilities. 🧵


I love the storytelling nature of @dwarkesh_sp in his podcasts How could I miss up on a book all about AI, where it was, where it is and where it will be This has me so excited! I have the audible now but wonder if I should wait till the book arrives 🙈

What is intelligence? What will it take to create AGI? What happens once we succeed? The Scaling Era: An Oral History of AI, 2019–2025 by @dwarkesh_sp and @g_leech_ explores the questions animating those at the frontier of AI research. It’s out today: press.stripe.com/scaling



This curve is only relevant until models become more capable @DevinAI in Early 2024 made so much noise, yet when everyone first used it, the reaction was meh Then came Sonnet 3.5, things suddenly changed The same shall happen again Prepare for this future

100% Maybe I'll eat my hat, but this is how I see it right now

thorstenball's tweet image. 100%

Maybe I'll eat my hat, but this is how I see it right now


I believe in the equilibrium of determinism vs non-determinism in LLMs to be owned by them Sonnet 4.5 is the first model to truly show emergent behaviour that correlates to it being able to steer its nature in either direction If you have any orchestrator or routing layer…


For anyone on the Cursor Ultra plan, how has it been compared to Claude Max and ChatGPT Pro subscription? Is there usage significantly lower or does it limit you compared to them?


The elixir of life and innovation

I want nothing other than time.



United States الاتجاهات

Loading...

Something went wrong.


Something went wrong.