Renjith

Co-Founder at Tweets are mostly 'read it later' bookmarks on Startups, Entrepreneurship, Psychology or Philosophy.

[Renjith @renjith_k]

The Only Way to Grow Huge http://t.co/DFshikmp1u

[Addison Huddy @addisonhuddy]

Just tested @litestreamio on @exedev — the pairing is perfect.

Built a sensor monitoring app with SQLite, set up continuous replication to S3, then ran a disaster recovery drill: deleted the database, restored from S3 in 2.3 seconds, lost only 5 seconds of data.

The whole stack — Go server, live dashboard, synthetic data generator, Litestream v0.5 with tiered compaction, S3 bucket — was built and deployed in one conversation with Shelley (https://t.co/4Y3KR1gtPp's coding agent). Systemd services, HTTPS, persistent disk, all just there.

[Renjith @renjith_k]

@E20Party #e20

[xanax @donttouchxanax]

An Indian institute filmed a professor teaching classical physics with no notes, no slides, and a green board. Almost nobody outside India watches it.

This is Classical Physics, lecture 1, taught by V. Balakrishnan at IIT Madras, published free by NPTEL, India's national open courseware programme.

Balakrishnan taught at IIT Madras for decades. Students who sat in that room describe the same thing: the whole subject held in his head, delivered in order, no paper on the podium.

Lecture 1 is the introduction. What classical mechanics claims, where the claims hold, and the exact points where they stop holding and quantum takes over.

Watch how he speaks. Complete sentences, no filler, each idea finished before the next begins. Rare in any lecture hall on any continent.

A physicist I know worked through the whole NPTEL series and rates it above the western courses he paid for.

Free on YouTube, 17 years online, low resolution, no marketing.

The teaching was never concentrated in the famous universities.

[Taelin @VictorTaelin]

every memory solution I tried sucks, so I rolled my own

it is a plug and play prompt that gives any agent infinite memory

db is an append-only log. compression on the spot (nap, don't sleep). nothing is ever deleted, memory context is constant-sized, details fade with age

[Renjith @renjith_k]

For readable thorough code, writing code with opus 5 medium and reviewing with sol high/xhigh may be one option.

[Renjith @renjith_k]

"It(Sol) won't do the sensible thing you forgot to write down. Its two known vices: over-engineering , and dense, hard-to-read code.

Opus 5 is the middle ground. Its code is noticeably simpler and more readable than sol's.
Its one weakness is a thoroughness gap."

[Renjith @renjith_k]

Review of Sol and Opus-5 by the big brother Fable:-

"Sol (gpt-5.6-sol) is the precision engine. Its strength is edge-case thoroughness.
It follows a spec literally, which cuts both ways: give it precise constraints and it honors every one; but spec boundaries become ceilings"

[Renjith @renjith_k]

@VictorTaelin As a heavy Claude Code user, have you felt this to be superior to Claude Code's memory implementation? Or in other words , how did you feel about Claude Code's memory implementation?

[Taelin @VictorTaelin]

I think these who fear "AI will take my jobs" ultimately fall to the following logical trap: "the value of my work might go down to zero (true), thus I'll not be able to afford the things that I could afford otherwise (false)".

Obviously, if the value of everyone's work goes to 0, then so does the *cost* of everything. Models are getting better at everything, so, there's no realistic scenario where YOU stop being valuable yet you still need to pay the same to all the goods and services you consume. But that is somehow a hard conclusion to make.

If I'd predict the future, this is how I think things will work out.

First, AI will take nearly all jobs that can be done behind a computer. Period. Stop coping. I'm glad I don't need to measure my words. That IS happening. Fable is almost implementing Bend on its own, in many instances it feels more competent than myself, and I've hardly met a single human I'd attribute that title to. Accept reality.

Second, once software and math becomes trivial and people get used to the fact they can conjure programs and proofs beyond human comprehension - once that becomes the new normal - science will turn its attention head to elsewhere. In particular, SF nerds will be pikachu-shocked to find out that solving math/swe doesn't automatically solve hunger, climate, violence, biology. So, these will become the next big thing in the frontier of science.

Third, and this is where most are worried about - what happens to our jobs? This discussion will become more and more prevalent and some think UBI will be the solution. I don't think so. I think UBI will not be *needed* because the cost of everything will naturally go down to zero.

I think this point is easy to visualize for things behind a screen. That utility app you pay a $20 / mo for? You will stop paying for it, because a local AI will rebuild the thing for $0.1, and someone will publish it for free. A 100% open App Store will probably emerge, and will be massively adopted because quality will be so high, and people don't want to pay for things they can get for free at higher quality.

This is "auto-UBI" for software. We all can afford it now, not because the gov is sending us a $5k check every month, but because the cost of the thing is zero. If you accept that, then, it shouldn't be much harder to accept this will ALSO happen, to a large extent, to things in the real world. The food you eat, the water you drink, the car you ride.

Real world goods cannot go down to literal zero because atoms aren't copiable, but their production will be so aggressively automated, each on their own way, that the practical result will be nearly the same. You'll walk into a store, grab stuff, and go home. You'll make a hand gesture, get into an auto driving car, and go wherever you want to.

And if you think that's not going to happen, that's probably because you have this "but who pays for it?" on your mind. And that's the catch, nobody does. AI robots are producing all the stuff needed to sustain it in such an enormous abundance that there is basically no cost. "What prevents vandalism and abuse?" - Laws don't stop working, they get better too. This is so hard to visualize because we were born to live in a world where you must work to *survive*. This is so axiomatically ingrained in our brains we can't imagine another way. Your children will not live in this world.

This is the future of humanity, and it will happen sooner than you expect. I'm not denying the transition won't be turbulent and I'm not invalidating these who are afraid. Some fields (swe for example) will be gone faster than others, and that IS a real concern. In particular, I think we should put a lot of pressure in ensuring this transition happens as smoothly as possible. But once it lands, humanity will be in a much better position than it is today. Your children will not work to live. They will just... live. And they'll look at work scales as view slavery today. Which is how it should be

[Renjith @renjith_k]

@JeffDean @henrydaubrez @OriolVinyalsML @quocleix Great timing to start recording! It was great to see a first person view from such close quarters, of the prime moment. Thought it was from iphone. Surprised to hear it was not.

[Jeff Dean @JeffDean]

Yesterday I was fortunate enough to go to my first-ever World Cup game, with my long-time colleagues/friends @OriolVinyalsML and @quocleix and their spouses). We were sitting in the corner area and had a quite good view of the goal!

I present to you a 43 second multi-part drama filled with emotion:

The initial promising-looking cross coming in, but looking overhit

Nico Williams cleverly knocking it down at the back post into a dangerous area

Ferran Torres striking it cleanly into the roof of the net

The crowd rising as one (forcing me to stand up as well)

The elation of the Spanish players racing off the bench to celebrate

The dejection of the Argentinian players, their defense having finally been breached in extra time

The elation of my Spanish colleague Oriol and his wife Meire next to me (he and I are both Barça fans, so it was nice to see a Barça player score the winning goal)

The entire stadium reacting

Whew!

[Mohamed Nowsath @Md_Nowsath_]

14 years of career, paid close to 1 cr as income tax so far. Now, being laid off.

Absolutely no trace of any support system. Thank you India 🙏🤐

[Skyroot Aerospace @SkyrootA]

Exactly 24 hours ago, a rocket lifted off.
Today, a billion+ dreams have taken flight. 🇮🇳🚀

Reliving the launch—from lift-off to satellite separation. Reaching orbit on the very first attempt of a new launch vehicle is an incredibly rare feat globally.

The world now has a new launch vehicle.

From India, for the world. 🚀

#Vikram1 #MissionAagaman

[Pawan @PawanKChandana]

It’s been 24 hours now…

Vikram-1’s launch still feels surreal.

Knowing just how hard this is, I couldn’t contain the excitement at Mission Control as the rocket we built with so many dreams successfully inserted satellites into orbit — on its very first attempt.

Just phenomenal! 🔥🚀🚀

[OpenShip @openshipio]

Introducing OpenShip.

An open-source application platform for building, deploying, operating, and scaling applications on infrastructure you own.

Replace deployment tools, managed services, and infrastructure workflows with one open-source platform.

Available today:

- Mail Server (Built in. One click).

• Runs on your own VPS
• Unlimited domains
• Unlimited mailboxes
• Modern webmail included
• Connect with Gmail, Outlook, Apple Mail, Thunderbird, or any IMAP/SMTP client
• Send email directly from your applications using SMTP
• No mailbox subscriptions or API fees
• High email deliverability

- Deployment:

• Deploy any stack
• Git-based deployments
• Zero-downtime deployments
• One-click rollbacks
• Development, Staging, and Production environments
• Multi-branch deployments with isolated environments • Deploy to VPSs, dedicated servers, cloud VMs, or your homelab

- Services:
Provision the services your applications need in one click.
Replace multiple managed providers with services running on your own infrastructure.

• Supabase
• PostgreSQL
• MySQL
• MariaDB
• MongoDB
• Redis
• MinIO
• Meilisearch
• Qdrant
• RabbitMQ
• Kafka
• ClickHouse
• Elasticsearch
...and deploy any other service alongside your applications.

- Operate:

• Live deployment logs
• Live request logs
• Real-time traffic analytics
• Automated backups
• Monitoring
• Secrets management
• Domains
• Automatic SSL
• Environment variables
• Scheduled jobs
• Health checks

-Multi Environments:

• Separate Development, Staging, and Production environments
• Deploy every branch independently
• Test changes before production
• Isolated services, secrets, and configuration per environment

- Security & Teams:

• Team management
• Role-based access control
• IP allow/block rules
• Rate limiting
• Security rules

- Developer Experience:

• Web dashboard
• Native desktop application
• CLI
• REST API
• MCP support for AI agents, just add the mcp and your agent can do the work for you
• Manage your infrastructure without living in SSH

- Coming Soon:

• Multi-server clustering for applications and databases
• One-click load balancing
• Horizontal scaling across multiple servers
• Built-in high availability and failover
• Scale from a single VPS to a cluster with the same workflow
• Just add servers. OpenShip handles the rest.
Open source.

[Skyroot Aerospace @SkyrootA]

Hello space, we have arrived! Vikram-1's Test Flight-1 has completed it's mission.

The first ever Indian private sector launch has been successfully completed.

#Vikram1 #MissionAagaman #OpeningSpaceForAll

[Renjith @renjith_k]

Gpt 5.6 models are more thorough, and catches more edge cases than any other model
Coupled with the auto compaction feature in codex, they can work on for days independently, given a goal.

[Renjith @renjith_k]

@onusoz Since Fable is about to go from subscription, what would be the best next?

[Ado @adocomplete]

The best SaaS platforms are ones that LLMs can effortlessly integrate into your app. Some recent winners for me:

- @resend - absolute pleasure to use
- @Cloudflare - Wrangler 🤝 Claude BFFs
- @ElevenLabs - effortless
- @WorkOS - best in class auth
- @PlanetScale - it just works

[Sam Hogan 🇺🇸 @samhogan]

We're releasing Inference AutoTune

Distill any frontier model into a 1-30B parameter task-specific SLM with only 25 lines of code

automatically route requests to reduce cost and latency by >90%

~2 hours and <$250 to train. You own the weights

Available in private beta today