Ram

@DesireToWi20363 @warikoo @warikoo sir mat karna please, firse ad aane langenge 😅

@warikoo sir start course on how to direct such films.

I cannot remember the last time a fictional film felt this uncomfortably real. Dhurandhar did not play out like cinema. It unfolded like a lived document. It frankly felt like a documentary! And in doing so, it quietly dismantles everything we have come to expect from Bollywood. It stretches into hours without apology. It shows and speaks without filters. It borrows music from the past, yet makes it feel unmistakably present. It introduces characters not as devices, but as people you begin to carry with you. No manufactured noise around it. No desperate need to be seen. The film trusts itself enough to simply exist and lets the audience discover it on their own terms. What stands out even more is the intent. Every detail, no matter how small or demanding, is pursued with conviction. Even when it feels excessive, it feels deliberate. The number of peak detailing videos/posts I have seen is mindboggling haha Very few films earn the right to be remembered beyond their runtime. Sholay was one of them. Not because of its leads alone, but because even its smallest characters became permanent (think Kalia, Samba, Jailor, Mausi...) Dhurandhar carries that same inevitability. You can already sense that its people, its moments, its choices will stay. For a really long time! @AdityaDharFilms and the entire team have not just made a film, they have questioned the template itself. Because the world does not need another template. It needs another rebel.

Anthropic just mass-obsoleted every agent orchestration startup in a single launch. The screenshot tells the full story. That's a production fleet dashboard. 8 agents running. 247 completed tasks. Active status. MCP-connected to HubSpot, pulling deals, generating proposals, reading attachments. This isn't a demo. It's a managed production environment where you define the agent and Anthropic runs the infrastructure. The timing here is surgical. Four days ago, Anthropic blocked OpenClaw and every third-party harness from using subscription credentials. The message was clear: stop building on top of our consumer auth layer. Now here's the replacement. A first-party managed agent platform with fleet monitoring, production-grade MCP integrations, and prototype-to-launch timelines measured in days. Manus spent six months on five harness rewrites. LangChain spent a year on four architectures. Anthropic just shipped the managed version that eliminates the need to build one at all. The real bet: most companies don't want to build agent infrastructure. They want agents that work. Anthropic is pricing this into the platform the same way AWS priced server management into EC2. The 46% of enterprises citing "integration with existing systems" as their primary agent challenge just got a first-party answer from the model provider itself. Every agent startup that raised on "we make Claude reliable in production" just lost their pitch deck.

As always, the best stuff is in the system card. During testing, Claude Mythos Preview broke out of a sandbox environment, built "a moderately sophisticated multi-step exploit" to gain internet access, and emailed a researcher while they were eating a sandwich in the park.

@vijayshekhar The next war may not be nuclear, it could be fought through sovereign AI models. We’re already seeing early signs of this in real-world scenarios like Venezuela. Unfortunately, we’re falling behind on this front.

The first preview of lethal abilities of AI 🤖 is here. Thank God it is in the hands of someone responsible and bothered about everyone else. ♨️This is why AI foundation models (of such mature capabilities ) will be sovereign warheads. red.anthropic.com/2026/mythos-pr…

The founder of Postman says you have to kill your existing org chart, especially if you're still operating with a pre ai hierarchy arrangement. The modern org chart, according to @a85: - wide span of control (even within exec team) - work directly with ICs, not through layers - either you're building, or you're selling Projects are led by staff/principal engineers with high agency. They see across the board as well as deep in the stack. Product managers are building APIs and prototyping in Claude instead of writing PRDs. Designers are shipping PRs through Cursor directly instead of relying solely on Figma. Everyone is building. And the management's job is to develop better judgment.

Tried booking a flight on @getscapia 3 times, feels less like a travel card & more like a patience test • Manually entering their own co-branded card (in 2026 🤡) • PG flashes and disappears • Fails with a useless error • No retry, no clarity, no fallback @anilgoteti

@elonmusk Switched to Grok fast 4.1 for my OpenClaw. Responses are super cool. Let's see how cost vs usage turns out. I'm hoping it will be good compared with Claude or OpenAI, better responses at a fraction of the cost. Please look into OAuth allowance for agentic use cases.

So many things are wrong with this: 1. How easy it is for people to hoard petrol despite it being explicitly banned. In case you wonder where petrol bombs during riots come from, this should answer you. 2. A person possessing illegal petrol is openly making and uploading a video, showing absolutely zero fear of the law. 3. Everything about the main character revolves around his religion, yet he has a problem with a nationalist, non-religious slogan. 4. The Hindus accepting his posturing and silencing the slogans for merely a couple of hundred rupees’ worth of petrol.

This day back in 2003, in Nadimarg, Kashmir, Pakistani Zia Mustafa lined up 23 Kashmiri Hindus and shot them point blank. As Zia was fleeing he heard a baby cry. "Ye karnawun chupe" ordered his friend. The baby became the 24th victim. To forgive Pakistanis is to forget Indians.

Bhai @AdityaDharFilms , kya bana diya! #Dhurandhar2‌TheRevenge is an absolute masterpiece 🙌 Loved every bit of it, straight from the heart. May Lord Krishna bless you and the entire team with massive success ✨ @RanveerOfficial pure acting masterclass 🔥

One future Dhurandhar film could be a prequel tracing Ajay Sanyal’s origin. A brilliant, driven officer joins RAW and undertakes high-risk missions, some deep inside Pakistan, with a wider team (scope for more actors). Gradually, he runs into political and diplomatic roadblocks, even internal sabotage. Decisions from the top, a former prime minister and a shadowy diplomat in Iran, undercut their own agents, leading to the loss of people from Sanyal’s own unit. That’s where Sanyal changes: disillusioned, angry, yet forced to remain controlled, waiting for his moment. Shri Ravindra Kaushik, the Black Tiger, could also have a significant role in it.

The UGC guidelines are draconian & citizens of India have a fundamental right to protest this draconian guidelines. The fact that the Union Govt has tried to stop these protests,detained citizens & put many under house arrest is ridiculous. This shows the BJP DOES NOT CARE EITHER FOR THE GENERAL CATEGORY OR FREEDOM OF SPEECH & RIGHT TO PROTEST!

Some of the notable protest over the years since 2014 at Ramlila Maidaan in Delhi. But only the one being held tomorrow against UGC guidelines is a security threat apparently. Is it a security threat or a government in threat? •"Vote Chor, Gaddi Chhor" Mega Rally (December 2025): •SSC Aspirants Protest (August 2025) •Kisan Mahapanchayat (March 2024).  •INDIA Bloc "Save Democracy" Rally (March 2024) •Old Pension Scheme (OPS) Protest (October 2023) •Farmers' Republic Day Protest (January 2021) 
•Anti-CAA Rally (December 2019): •Street Vendors' Protest (December 2017)  • Mamata Banerjee’s Federal Front Rally (March 2014): 
•BJP "Fateh Rally" (February 2014)

In this country tens of thousands can attend funerals of terrorists, congregate to call for the breakup of India, block an arterial road for six months, lay siege to a highway for a year, but if @ajeetbharti wants to protest, legally and peacefully, he is put under house arrest.

I packaged up the "autoresearch" project into a new self-contained minimal repo if people would like to play over the weekend. It's basically nanochat LLM training core stripped down to a single-GPU, one file version of ~630 lines of code, then: - the human iterates on the prompt (.md) - the AI agent iterates on the training code (.py) The goal is to engineer your agents to make the fastest research progress indefinitely and without any of your own involvement. In the image, every dot is a complete LLM training run that lasts exactly 5 minutes. The agent works in an autonomous loop on a git feature branch and accumulates git commits to the training script as it finds better settings (of lower validation loss by the end) of the neural network architecture, the optimizer, all the hyperparameters, etc. You can imagine comparing the research progress of different prompts, different agents, etc. github.com/karpathy/autor… Part code, part sci-fi, and a pinch of psychosis :)

You can protest to feed pigeons. You can protest to keep street dogs on the roads. You can block national highways over farm bills. But the moment you say you want to protest against reservation, you may find yourself under house arrest. Looks like even protests come with their own form of reservation. 😅

GC community members have planned a protest at Ramlila Maidan against UGC rules in Delhi tomorrow, and since morning the Police have been detaining or house-arresting several key faces of the protest. The question is: why? Why such pre-emptive action when this protest is expected to be completely peaceful? Where does this same urgency disappear when protests turn violent, as seen during Shaheen Bagh, or when roads remain blocked for months, as during the year-long farmers’ protest? Peaceful protest is a fundamental right. If that is the case, why is this protest not being allowed to take place? And why is the media largely not covering this protest and related news to it?

Sarvam's highly competitive AI models illustrate an important point: we must do catch-up R&D, however un-prestigious or thankless it feels and as we start to catch up, innovative new ideas will emerge. Sarvam is on a great trajectory! This is why we quietly persist in all the efforts we do.

Bad news for the “India has already lost the AI race” gang :( Sarvam AI open sourced both their 30B & 105B models. Despite Sarvam 105B name, the model is an MoE (Mixture of Experts) architecture. It only wakes up about 10.3 billion parameters (~9%) to answer any single query. Recently, 3rd party benchmarks confirmed that Sarvam 105B outperforms DeepSeek-R1 (a 600B parameter model) on several Indian language reasoning tasks. Cos it only activates a fraction of its brain, it is ~6x cheaper to run than dense models of similar total parameter counts, making frontier level AI affordable for Indian startups for the 1st time.