Bhaulik Patel
@deployengineer
Forward Deployed Engineer; sharing my AI findings; podcast host @ deployed engineer podcast
every LLM would be at least 10% better if they augmented current date just at the start of the conversation
Highly recommend watching @dwarkesh_sp and @karpathy podcast. Learnt a lot about limitations of the current LLMs and the current context around Nano GPT
The @karpathy interview 0:00:00 – AGI is still a decade away 0:30:33 – LLM cognitive deficits 0:40:53 – RL is terrible 0:50:26 – How do humans learn? 1:07:13 – AGI will blend into 2% GDP growth 1:18:24 – ASI 1:33:38 – Evolution of intelligence & culture 1:43:43 - Why self…
the real metric for the success of an AI product is how much double checking do users have to do each time they use your product
when internet happened, information was available through air, in the AI world, intelligence is available through air
Excited to ship Claude Haiku 4.5 today! What was state-of-the-art 5 months ago (Sonnet 4) is now available at 1/3 the cost and 2x the speed. Even beats Sonnet 4 at computer use. Available today wherever you get your Claude :)
Excited to release new repo: nanochat! (it's among the most unhinged I've written). Unlike my earlier similar repo nanoGPT which only covered pretraining, nanochat is a minimal, from scratch, full-stack training/inference pipeline of a simple ChatGPT clone in a single,…
folks saying they don’t need figma because they can use AI lack taste
i noticed a weird phenomenon, the pre-ai super star developers are finding every reason to justify why AI is a bubble. I'm not against them but its just a pattern i've noticed. If most of them just dabbled with this tech, they would realize how useful it is despite its limits
What the fuck just happened 🤯 Stanford just made fine-tuning irrelevant with a single paper. It’s called Agentic Context Engineering (ACE) and it proves you can make models smarter without touching a single weight. Instead of retraining, ACE evolves the context itself. The…
Life punishes the vague wish and rewards the specific ask. After all, conscious thinking is largely asking and answering questions in your own head. If you want confusion and heartache, ask vague questions. If you want uncommon clarity and results, ask uncommonly clear questions.
the stealth cheetah model on Cursor gives me the same level of outputs a Claude Sonnet 3.5 except faster
This is how Anthropic decides what to build next—and it's brilliant. Instead of endless spec documents and roadmap debates, the Claude Code team has cracked the code on feature prioritization: prototype first, decide later. Here's their process (shared by Catherine Wu, Product…
AI engineering is the act of piping static systems into more intelligent ones
United States Trends
- 1. #SmackDown 22.1K posts
- 2. #WorldSeries 69.3K posts
- 3. Snell 8,005 posts
- 4. Cole Anthony 1,483 posts
- 5. Halo 138K posts
- 6. Celtics 18.3K posts
- 7. Darryn Peterson 1,723 posts
- 8. #TheLastDriveIn 1,207 posts
- 9. Jade Cargill 4,938 posts
- 10. Knicks 22K posts
- 11. PlayStation 64.9K posts
- 12. Grizzlies 3,629 posts
- 13. Purdue 6,423 posts
- 14. Yesavage 7,274 posts
- 15. Paolo 11K posts
- 16. Kyshawn George N/A
- 17. Zelina 1,591 posts
- 18. #LetsGoDodgers 6,125 posts
- 19. Xbox 78.3K posts
- 20. Zion 16K posts
Something went wrong.
Something went wrong.