Suresh Ramkrishnaiah

Engineer | Now onto AI | Ex-bunch of random startups |

[Suresh Ramkrishnaiah @sureshgowda]

Easy Local OpenClaw setup for learning.

(1/n)

System Requirements:
I have been able to setup and run this on 10 year old PC with 16GB RAM, 4-Core CPU.

Goto https://t.co/kODxc8f39C and run the one-line command which will walk you through the setup process.

[Suresh Ramkrishnaiah @sureshgowda]

@ryancarson I disagree. Build your own AI factory, harness, use models based on the task and the factory must reflect the organization and products you are building. If you don't you will have a generic factory and tied to the vendor (quality and cost).

[Suresh Ramkrishnaiah @sureshgowda]

@jomatech That is awesome man! Storytelling is great!

[Suresh Ramkrishnaiah @sureshgowda]

@vasantshetty81 @mundhebanni He is onto something big actually!!!

[Suresh Ramkrishnaiah @sureshgowda]

@Steve_Yegge @superlogical Lot of terminal wars lately. I am trying wezterm, alacritty, herdr, tmux and none of them work 100% seemlessly. Whoever solves both mouse/keyboard actions without going through numerous config hacks is going to win the terminal war.

[Emm | scenario.com @emmanuel_2m]

I'm not sending my deck anymore 😂

[Suresh Ramkrishnaiah @sureshgowda]

I agree, Also this is a common pattern I am noticing across tasks. Find standards and apply that to the tasks. Design standards for design tasks, engineering standards for engineering tasks etc.

[Kun Chen @kunchenguid]

people who laugh and say karpathy's latest tips are outdated - i suspect the vast majority of them have not even tried the tips yet

i just tested the ASD-STE100 wording rule and it's surprisingly good at helping increase clarity of model response, even within html artifacts. but the trick is that the full ruleset is a bit too strict and you need to pick a subset

one prompt you can run super easily:

"randomly sample 10 session transcripts where i worked with you interactively within the past week. apply ASD-STE100 rules to assistant responses and analyze which rules would have increased clarity, reduced confusion and improved the conversations, then document those rules in my user level AGENTS.md"

you might be impressed!

[Suresh Ramkrishnaiah @sureshgowda]

That is "monster" upgrade! 😂

[Sundar Pichai @sundarpichai]

Importantly Argon has frontier safeguards and we are rolling it out responsibly - it’s with the US gov’t and going to a set of trusted cyber defenders through our Fairwind Program today.

We’re going to make it available as soon as we can and as safely as we can. So hold tight, lots more coming, and you’re going to see us iterating rapidly.

https://t.co/8MiXIfXIKr

[Sundar Pichai @sundarpichai]

Importantly Argon has frontier safeguards and we are rolling it out responsibly - it’s with the US gov’t and going to a set of trusted cyber defenders through our Fairwind Program today.

We’re going to make it available as soon as we can and as safely as we can. So hold tight, lots more coming, and you’re going to see us iterating rapidly.

https://t.co/8MiXIfXIKr

[Suresh Ramkrishnaiah @sureshgowda]

@ARmastrangelo @GE_Appliances Yes, experienced the same.

[Suresh Ramkrishnaiah @sureshgowda]

@GergelyOrosz "We are writing deterministic code from non-deterministic agent" - An engineer can never trust that. It's not engineering it's science.

[Suresh Ramkrishnaiah @sureshgowda]

@alpasankhyata There are few that you play and this one is a must: https://t.co/0Xle8HNcup

[Suresh Ramkrishnaiah @sureshgowda]

@larrykim Basically back to fundamentals (math and physics)! and then go live like a caveman! 😂

[Suresh Ramkrishnaiah @sureshgowda]

Keeping pace with frontier AI knowledge https://t.co/FdwRdkPtJn

[Suresh Ramkrishnaiah @sureshgowda]

Muse Spark 1.3 release. Use muse-spark-1.3-contributor to share prompts for lower token cost. Is this a new strategy to make choice explicit?

[Suresh Ramkrishnaiah @sureshgowda]

Nice! Available to public!

[Mihail Eric @mihail_eric]

I’m excited to finally announce the newest edition my Stanford course 𝗧𝗵𝗲 𝗠𝗼𝗱𝗲𝗿𝗻 𝗦𝗼𝗳𝘁𝘄𝗮𝗿𝗲 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗿. It has been 9 months in the making.

Last November, with the release of Claude Opus 4.5, coding agents experienced a step function improvement in capability. We all felt it. The LLMs were more powerful, could reason for longer, solve harder tasks.

This year’s iteration of my course reflects the 2026 metamorphosis of software engineering.

My core belief is simple: AI-native developers of the LLM era are going to become the most important members of any software organization. I have designed my course to train this next generation of engineers.

𝗪𝗵𝗮𝘁’𝘀 𝗱𝗶𝗳𝗳𝗲𝗿𝗲𝗻𝘁 𝘁𝗵𝗶𝘀 𝘁𝗶𝗺𝗲 𝗮𝗿𝗼𝘂𝗻𝗱

First, 85% of my Fall 2025 class material is being thrown out. The Fall 2026 syllabus reflects the core capabilities AI-native engineers must have: agent skills, advanced context engineering, MCP portals, agent-ready codebase principles, agentic code review, security, parallelizing background agents, software factories, and more.

Second, I am going to teach my students how to have software taste. Every student will be required to ship pull requests to production-grade, real-world codebases. The course is collaborating with the top open-source AI repos who will offer support and mentorship to students on how to meaningfully contribute to their projects.

This has never been done before in any university course so I am incredibly grateful to our OSS Partners: @browserbase, @HeyGen, @CopilotKit, @semgrep, @OpenHandsDev, @milvusio, @marimo_io, Pi, @crewAIInc, @warpdotdev, @vercel, @cmux, @arizeai, @UnslothAI, and @anyscalecompute.

𝗪𝗵𝗮𝘁’𝘀 𝘀𝘁𝗮𝘆𝗶𝗻𝗴 𝘁𝗵𝗲 𝘀𝗮𝗺𝗲

I’m fortunate to again have AI software engineering leaders and founders as guest speakers to share their learnings from building top coding agent products. Thank you to @leerob from @cursor_ai, @bcherny of @claudeai code, @EnoReyes of @FactoryAI, @silasalberti of @cognition, @0xine of @semgrep, Rajesh Bhatia of @Cloudflare , @amasad of @Replit, and @eladgil.

All resources will be available online. All classes will be available to the public.

9/22 on Stanford campus. See you in class. 

https://t.co/wTokHyUMsz

[Suresh Ramkrishnaiah @sureshgowda]

@garrytan @ycombinator They are already… deterministic harness. You can make them dynamic by adding probabilistic capabilities.

[Suresh Ramkrishnaiah @sureshgowda]

If we can generate lot of tokens at high speed, then we can solve most of the problems with brute force algorithm compared to a human using most efficient problem solving technique as long as the total time is lower than human trial. Token Quality = Efficiency.

[Suresh Ramkrishnaiah @sureshgowda]

@pixawna Inspiring story. Hope you become an entrepreneur one day!

[Suresh Ramkrishnaiah @sureshgowda]

@petergyang @tobi Being Apple product user for years this issue is the most frustrating among all! Will try it. Thanks