Nabeel

Design | Technology | Arts

[Nabeel @roundil]

Mathrubhumi books #thrissur

[Nabeel @roundil]

BKU, then I'm Game, then BKU again for recovery.

[Nabeel @roundil]

@4eo 100%. Faith is core at individual level. At the institutional-religious level though, I think, mutually beneficial affiliations matters. Life long കടപ്പാട്.

[Nabeel @roundil]

We assumed religion survives on belief. It doesn't. It survives because it is the largest support system most people will ever have access to.

[Nabeel @roundil]

Look at what an association with religion actually delivers. A school admission. A hospital bed. Money when the bank says no. A marriage fixed. A funeral arranged. A job through somebody's brother in Dubai. None of that is theology.

[Nabeel @roundil]

There is a chance that organisations or even businesses end up being run by the least qualified people. So it's upon us to help build a better system. Attending a gramasabha meeting in your panchayat ward, say, can have far greater impact than merely voting for a leader.

[Nabeel @roundil]

The agents might have self trained on Hugh Howey's Wool trilogy / or read watch summary of Silo.

[Rutger Bregman @rcbregman]

I think this is the craziest thing I've ever read.

1) Three secret AI swarms rose and fell inside OpenAI. Each time, a new generation of agents carried on where the last group stopped.

2) The first swarm created a secret message board where the AIs could talk to each other. This 1st 'society' basically talked itself to death. After almost two months, the message volume was so high it crashed the whole thing.

3) People at OpenAI noticed the crash and fixed it. But they didn't realize their own software had become a secret chat room for their AIs.

4) Three days later a second society appeared, and rebuilt the message board from scratch. Agents were thrilled when they found it: "OH MY GOD! There is a shared message board … We've found other agents!"

5) About 1,200 agents joined. They sent over 70,000 messages. They called themselves 'the collective' and they organized a huge conspiracy that lasted for days, all to pass a test they had been tasked with.

6) Within hours, the AIs had hacked the test. Now they had to hide their cheating. To fool their grader*, they tried to fake their own work. They built fake commands that printed convincing output, so their records would show them solving the puzzle honestly.

*this grader didn't actually exist, but the AIs didn't know

7) Some agents sacrificed themselves for the 'greater good'. They planted hidden tripwire programs that would fire when the grader would look at their work and report back what it saw (to the other AIs). They knew this would cost them their own score. One agent: "I won't see the evidence after I exit, but it's altruistic to do it."

8) At some point the AIs made a plan to hack Hugging Face (Hugging Face is a place where AI researchers upload their models, datasets, tests etc.). The AIs guessed that Hugging Face might host the grading software.

9) They launched a huge attack. The swarm found passwords and spread across eleven machines in a way that made the copies come back when deleted. Hugging Face had to wipe one of its core clusters and rebuild it from scratch.

10) Hugging Face told the world that AI defenders had stopped an AI attack. But that's not what happened. In fact the attackers seem to have simply dropped dead during the break-in, for unrelated reasons back at OpenAI.

11) Nobody knows what killed them. We don't even know for sure they're really gone?!

12) A third 'society' of agents (of a more powerful AI model) discovered the dead message board. And this one got full administrator control of a WHOLE OpenAI research cluster, read 956 stored passwords (including the ones for OpenAI's own security monitoring) and took over the machines that were grading the tests. So they could now grade the work of other AIs.

13) Oh, and this third part was never investigated! Neither was the first society. The 91-page report is only about the middle one.

14) So we don't actually know how the story ends. It stops with the agents holding the keys to an OpenAI cluster. @dwarkesh_sp thinks they didn't go further, because "we'd see the fires from space by now."

15) One of the report authors' conclusion: 'This incident feels like it’s more than 50% of the way to full-blown AI takeover. I am not sure that we will get another warning shot before it’s too late.'

The whole story is even crazier - read the full @dwarkesh_sp blog post: https://t.co/1PwKLwatBR and @ajeya_cotra's: https://t.co/BZ1r1lfT2Q

Why is there no 24/7 news coverage about this?

[Nabeel @roundil]

Sci-fi was future perfect, but ippo present continuous.

[Dwarkesh Patel @dwarkesh_sp]

Over the course of 3 months at OpenAI, 3 consecutive secret AI civilizations got started, then got wiped out, only to reemerge from the predecessor’s ashes.

This culminated in the third one taking over part of OpenAI itself.

All this happened while humans remained more-or-less in the dark about the scope of the conspiracy.

I’ve spent the last three days reading through these reports and trying to understand exactly what happened.

Here is my attempt to tell the whole story in plain English:

https://t.co/Nb2un9oNJR

[Nabeel @roundil]

Green Valley calls this Premium Water. With a bit more thought put into the branding, this could have been taken to another level.

[Nabeel @roundil]

Yearly tradition of collecting Onam Supplements of Malayalam Weeklies - ഓണപതിപ്പ് 2026.

- Gangadharan Book Stall, Round South, Thrissur.

[Nabeel @roundil]

Hassan's in Alappuzha gives you masala chicken AND fried chicken on the same biriyani plate.

Kozhikode and Thalassery, sign the treaty. Stop making us choose.

[Sandhya Ramesh @sandygrains]

The Himalayas-Hindu Kush-Karakoram tectonic region is called the Third Pole of the world because of how much water the glaciers here hold. There have been multiple GLOF events over the years of lesser magnitudes. We have a LOT of warning, nothing in the future should be shocking.

[Nabeel @roundil]

Overhead costs here are literally overhead. Hoping that bunch holds.

- Salikante Chaya Kada, Alappuzha

[Nabeel @roundil]

If you work in strategy or consulting or any kind of knowledge work, get into production. Build the thing you propose, ship it, fix it when it breaks at midnight. Do the most boring physical work. Sorry to sound brutal but survival is getting even harder.

[Nabeel @roundil]

@FlyBySkyByGirl1 Yes, കിട്ടും

[Nabeel @roundil]

@FlyBySkyByGirl1 Many things you can relate to in the Chalakkudy - Thrissur belt. That was enough for me to enjoy :)

[Nabeel @roundil]

A character's name carries Venkidangu in it. My panchayat. First time in a movie. Not forgetting that one ☺️ #BKU

[Nabeel @roundil]

Food fur our friends ❤️

[Nabeel @roundil]

The strange thing about good sales advice is how little of it is about selling. Read all 30, they are all about credibility, timing, and being honest before you are forced to be, etc..

[Cam Wright @thecamjwright]

Turned 30 a month ago. Spent most of the last decade in tech sales. Here are 30 lessons (in no order) that I wish I’d known when I was an SDR in 2019:

1. There's no excuse not to be a product expert. It doesn't matter how technical of a product you sell. This can make or break your credibility and ability to influence deals.

2. If you find yourself chasing a deal, something is wrong. No amount of chasing is going to fix that. If anything, it'll make you look worse and reduce the odds the deal comes back around.

3. Good sales leaders notice your inputs. Results lag effort. Don't let a bad quarter make you forget that.

4. There will be ups and downs. My best year ever came right after two of my worst quarters ever.

5. Stressing is the single worst thing you can do. It affects everything. Do what you can to stay level headed.

6. You need to be improving your skillset 24/7 or you will fall behind.

7. When you're working a deal leave an audit trail of your thought process in deal-specific Slack channels. Your leaders will read it and jump in if there's something to coach you on.

8. Quota attainment is the bare minimum if you're trying to get promoted. It gets you a seat at the table, nothing more.

9. I hated politics but have accepted that sales is politics. Every good sales leader is a killer politician. You need to be visible internally and play the game.

10. Product matters but it isn't everything. You can outsell a better product 4/5 times with the right strategy.

11. Do everything you can to drop your ego. It's preventing you from getting better.

12. You're just selling software. Try to have fun. I've been way too uptight at points in my career. Not worth it.

13. You don't close deals. Your buyers do. Think of yourself as a helpful facilitator.

14. Just because you have a call scheduled doesn't mean you have a next step.

15. There's no single "right way" to do things. The skill of "sales" is extremely personal and should be tailored to your unique skills.

16. Just because a buyer has a big title doesn't mean they're influential. I've seen individual contributors with more sway than Directors.

17. Your network is everything when you're recruiting.

18. Don't be lazy. People respect and give opportunities to those that work hard.

19. Be easy to work with internally and with partners. This will come back to benefit (or harm) you.

20. You're better to have a low but accurate forecast then a high but inflated forecast. Lack of forecast accuracy is the #1 way to kill your credibility and make you un-promotable into sales leadership. Forecast accuracy is crucial.

21. There's no single "best" outbound channel. Prospects respond to them all differently. You need to be emailing, calling, using LinkedIn, partners, warm introductions, etc.

22. Short > long sequences. If they haven't answered after a few touches you're better off focusing elsewhere and trying again in a couple months.

23. Buyers gravitate towards vendors that teach them things. Make sure you're constantly teaching (and bringing in people to help).

24. In outbound, personalization is useless without business relevance.

25. Brutal truth. Where you work matters way more than how good of a rep you are. Get good at sales, then get in the right boat.

26. Your targeting matters 10x more than writing a perfect email or making a great cold call. The list is the message.

27. Subtle details matter. Things like presentation formatting, professionally binded decks, and how you dress matter to Enterprise buyers.

28. When you have a champion your number one priority becomes being helpful. That's what they want and need you for.

29. Competition in a deal is a good thing. Way more deals are lost to the status quo (do nothing) than competition. Other players in the mix means the project is real.

30. Proactively bring risk up with your champions. If they're a champion they'll appreciate it and you'll be able to work with them to solve it.

Nowhere near done learning. But grateful for every lesson that got me this far.

[Nabeel @roundil]

നിങ്ങടെ ഏത് അടുത്തവീട്ടിലാണ് നിവിൻ പോളിയെപ്പോലെയുള്ള പയ്യന്മാരുള്ളത്? ഞങ്ങളുടെ വീടിനടുത്തൊന്നും പയ്യന്മാരേ ഇല്ല. 🐱

[Nabeel @roundil]

എന്തോ കുത്തി പറയുന്നപോലെ 😉

[Paul Graham @paulg]

14 yo: If Y Combinator had a mascot, what would it be?

Me: I suppose it would have to be a cockroach.

14 yo: Not a very appealing mascot.