Saliya Ekanayake

Some conversations make it obvious we’re in a once-in-a-generation shift in AI. @sidsheth shared his take on the watershed moments that brought us into the #AgeOfInference — and why breaking the memory & bandwidth wall now defines all future AI infrastructure. Great to dig into it with Gavin Baker & Mark Wade… and yes, we all agreed @elonmusk’s idea of data centers in space actually makes a lot of sense. 😉 #AI #GenerativeAI #dMatrix

When I created Claude Code as a side project back in September 2024, I had no idea it would grow to be what it is today. It is humbling to see how Claude Code has become a core dev tool for so many engineers, how enthusiastic the community is, and how people are using it for all sorts of things from coding, to devops, to research, to non-technical use cases. This technology is alien and magical, and it makes it so much easier for people to build and create. Increasingly, code is no longer the bottleneck. A year ago, Claude struggled to generate bash commands without escaping issues. It worked for seconds or minutes at a time. We saw early signs that it may become broadly useful for coding one day. Fast forward to today. In the last thirty days, I landed 259 PRs -- 497 commits, 40k lines added, 38k lines removed. Every single line was written by Claude Code + Opus 4.5. Claude consistently runs for minutes, hours, and days at a time (using Stop hooks). Software engineering is changing, and we are entering a new period in coding history. And we're still just getting started..

Nothing beats developer productivity than a desktop under your desk :)

@JeffDean @herron_brandan @Waymo The play on words is the best here : 😁

@herron_brandan @Waymo We'll cross that bridge when we come to it!

Exciting expansion! @Waymo now serves the whole SF Bay Area Peninsula from SF to San Jose and is taking riders on freeways. waymo.com/blog/2025/11/t…

Lyrics I had written sometime back comes to life with Suno.ai. Enjoy! Pata Aran Akase | පාට අරන් ආකාසේ .... youtu.be/w1K-FEhcIW0?si… via @YouTube

@sanjiva Interesting!

Huh .. I was wondering why A4 has an "odd" size of 210mm x 297mm .. and learnt interesting facts! Ratio of the two sides is 1 : sqrt(2). That was chosen because then doubling or halving maintains the ratio. The "odd" size is because A0 was defined to be 1 sqm. Then maintaining the ratio and dividing defines this size. Longer explanation from ChatGPT: chatgpt.com/share/68e3401c…. So cool! It also notes that US Letter (8.5x11in) is pretty much historical accident and not done with a geometric goal.

He loves attention so much!

@karpathy This could very well be an AI agent

Every company needs a DM POC - someone high up who you can just DM the most obvious things and who shortcuts the PM hierarchy.

@SunoMusic Love your product. One feature I am really missing is the ability to enforce a tune. For example during generation I'd find a particular one is nice. Then I'd want to use it with maybe slightly modified lyrics. Doing this today is extremely hard

in three words or less, what do you want to see from us in the future?

@elonmusk @elonmusk "no GPUs were harmed in the making of this" 😃

No AI was used to create this

@rasbt @ManningBooks Funny, I was searching for a tutorial on reasoning models and this post showed up timely. Pre-ordered!

If you want to check it out, it's available for pre-order @ManningBooks here: mng.bz/Dwra You will get immediate access to the first chapters, each new chapter as it's released, and the full book once complete.

I’ve been working on something new: 📚 Build a Reasoning Model (From Scratch). The first chapters just went live! (The book will cover topics from inference-time scaling to reinforcement learning)

@sameerajayasoma Love the logo

I recently had the opportunity to present OpenChoreo at WSO2Con Asia 2025 in Colombo, Sri Lanka. In this talk, I shared how the project got started, where it’s headed, and how it works, with a live demo of OpenChoreo and its integration with Backstage. Had to improvise a bit at the beginning while waiting for the slides to show up 😅 If you’re curious about internal developer platforms or want to see what we’ve been building, feel free to check out the recording below. youtube.com/watch?v=9A-CIy…

#GrokImagine is absolutely mind-blowing for both the quality and speed of generation. Almost scrollable speed

@elonmusk Time to build a X phone? I phone is not improving for years anyway

Why are the Grok and the 𝕏 app excluded from every list, except those measuring raw downloads, but ChatGPT is on every list? This is messed up!

Kohl's has just become an Amazon return center only 🙂

@rasbt Stay strong, wishing you speedy recovery

Usually, I post AI content and not personal life updates, but it somehow helps to get this off my chest. If I haven't been responding to messages, emails, or PRs, it's not because I don't care or don't want to. It's just physically hard at the moment, even with aids like speech-to-text. Life can be brutal sometimes. I was just getting back on my feet in March after a back injury that took 3 months to heal. I was so excited to be back coding and writing. And I had so many plans for this year. AI is a fast-moving field! And most importantly, I was hoping for this to be my last ER and hospital visit in life (or at least the next few decades), but no, life had other plans, and I have been dealing with a bad neck injury for the past 3 weeks. I have been seeing multiple doctors to get multiple opinions, and unfortunately, it may take 1–6 months until I can expect an improvement and start healing phase. So many talks and projects to cancel. And the worst feeling is the fear of missing out. Missing out on life and doing normal things like going for a hike on a beautiful spring day. And missing out on all the exciting AI developments and the missed project opportunities. I am working hard to get back on my feet as soon as possible or at least do some things even if it means I can do less and have to do things more slowly. I have done it before and can do it again!

@JeffDean This is cool. @JeffDean when could we expect Gemini to be integrated with Android auto? I'm already a happy paying customer and yesterday while driving had to figure out something and I was like ah, wish I could just ask from auto

"Write an implementation of this paper using Tensorflow.js and visualize the training with Canvas." Thanks, Gemini!

@ideogram_ai this is mind blowing. I've never seen this level of quality with generated text embedded images

@JeffDean I'm a happy paying customer of Gemini pro. Loving the integration. I'd love to see Gemini getting integrated to Gboard in addition to specific applications.

Our Gemini 2.5 Pro model has made significant improvements over the Gemini 2.0 series. It's nice to see it topping the LiveBench leaderboard by a pretty healthy margin (+6 in overall average score, +~10 in math and data analysis categories, +~2.5 in language) livebench.ai

New paper: turns out you can train deep nets without normalization layers by replacing them with a parameterized tanh()