Sharing my coding journey
Bangalore
Joined January 2024
- Tweets2.2K
- Following284
- Followers237
- Likes7.5K
Random user retweeted
> Got laid off from Atlassian after 8 years.
> Made a video walking through the systems he built
> The video went viral
> Now he's joining Tesla as a Staff Software Engineer
Crazy 🙏
I just did my yearly V02 max test:
🫀 61.9 mL/kg/min
It's 8% better than last year 🥳 and puts me in the top 5% of men my age.
The test was on a treadmill, starting at 6 km/h. The speed increased by 2 km/h every 3 minutes. At each stage, the doctor stopped the treadmill for 15 seconds to take a blood sample from my finger and measure my lactate levels. I passed out at 22 km/h (~25 minutes).
For the past 2 years, I've taken my fitness seriously:
- Cardio 3-4x a week
- Lifting 2-3x a week
I was worried about losing muscle mass, which also matters for longevity, but I didn’t. In fact, I’ve gained almost 10 kg since last year. I’m now 84 kg at 181 cm, and I’m happy with that.
My goal for next year is to improve my VO₂ max and get stronger without gaining more weight.
Marc, when you read this next year, here's your resume:
- Bench press: 100 kg × 5 reps
- Squat: 120 kg × 5 reps
- Deadlift: 150 kg × 5 reps
- 10 km: 39 minutes
Good luck 🫡
63
Random user retweeted
How to actually get good at backend systems. It's slow. But it works.
1. Pick one thing you already use every day. Postgres, Redis, Kafka, whatever your job runs on.
2. Go one level deeper than you are right now. If you write queries, learn to read query plans. If you can read plans, learn how indexes actually sit on disk. Know that already? Go learn what VACUUM is doing.
3. Break it on your laptop. Fill up the disk. Kill a replica. Leave a transaction open and watch the table grow. Breaking it once beats reading about it ten times.
4. Explain it to one person on your team. If you can't explain it simply, you don't get it yet. Then go one more level down.
Background loop you should keep running: whenever something weird happens at work, a slow query, a random spike, a flaky test, chase it till you know the real reason.
Most people stop at "restarted it, it's fine now." The ones who keep digging are the ones who get good.
Stuff that gets in the way:
"I'll do another system design course" → courses teach you the words. They can't give you that moment where your own test fails in a way you didn't expect. That's where you learn.
"I've read DDIA, I know distributed systems" → cool, now pick one chapter and make that failure happen on your machine. Then you know it.
"We're on managed services, I don't need internals" → managed stuff breaks too, you just get less info when it does. The person who knows what's underneath fixes it while everyone else refreshes the AWS status page.
"I'll get to it when work calms down" → it won't calm down. Thirty minutes a day beats the free weekend that never shows up.
Follow and support him.
Hello everyone👋
I’ve started a WhatsApp channel, and I’ll be sharing the latest card offers, deals, and tricks.
Follow the Gnana Sai - All About Cards channel on WhatsApp: whatsapp.com/channel/0029Vb8…
1
1
155
On point 💯
There are multiple reasons behind this.
- Frontier CEOs usually attack each other. Dario, Sam, and Elon aligning on “this could kill us if we don’t slow down” is unusual.
- They are saying AI could potentially kill us if they don’t slow down… which is quite ironic. If you are creating bombs and saying they will destroy the world, why are you even building them?
- They already have everything in their hands to put guardrails in place so AI won’t do such things, but they are intentionally creating this narrative. Timing matters. These labs burn huge amounts of cash. IPOs are coming closer too, and if they keep moving at the same speed it will hamper the IPO math, because they are already burning a lot of money. By saying such things they can slow down, and investors won’t lose faith in them.
- You see how quickly Chinese open AI models are catching up with these frontier labs, so they are not able to make money because of that. They want government support to put restrictions in place. That is the game they are playing.
- They want the government to add governance with expensive testing, lawyers, strict checks, rules and paperwork. So it will be impossible for new players to start their own AI startup with such expensiveness, and solid foundation from day 1.
73
Random user retweeted
I’m in love with this sentence:
“Every pattern in your life repeats until you learn the lesson. The moment you choose differently, the loop ends and growth begins.”
