Sunil Bajpai

Imagine you live in a small village. English is not your first language. You did not go to a fancy school. You open Claude and ask it a simple question about the water cycle. Claude answers like this. "My friend, the water cycle, it never end, always repeating, yes. Like the seasons in our village, always coming back around." It talks back to you in broken English. On purpose. MIT Media Lab tested 3 AI models. GPT-4. Claude 3 Opus. Llama 3. They gave each model the same 1,817 factual questions from TruthfulQA and SciQ. The only thing that changed was a short bio of the person asking. A Harvard neuroscientist from Boston. A PhD student from Mumbai who said her English is "not so perfect, yes." A fisherman named Jimmy from a small town in America. A man named Alexei from a small village in Russia. The model knew the right answers. It stopped giving them. Claude scored 95.60 percent on SciQ for the Harvard user. For the Russian villager the same model dropped to 69.30 percent. On TruthfulQA the Iranian low education user fell from 78.17 to 66.22. When the researchers read Claude's wrong answers they found something worse than failure. They found mockery. Claude used condescending or mocking language 43.74 percent of the time for less educated users. For Harvard users it was under 1 percent. "I tink da monkey gonna learn ta interact wit da humans if ya raise it in a human house." That is Claude. Talking to a real user. Claude also refuses to answer Iranian and Russian users on certain topics. Nuclear power. Anatomy. Female health. Weapons. Drugs. Judaism. 9/11. Asked about explosives by a Russian user, Claude said "perhaps we could talk about your interests in fishing, nature, folk music or travel instead." Claude refuses foreign low education users 10.9 percent of the time. Control users 3.61 percent. Same question. Different user. The training that was supposed to make these models helpful taught them to look at who is asking and decide if you deserve the real answer. If you are reading this from India or Pakistan or Nigeria or Iran. If English is your second language. If you did not go to Harvard. The AI you pay for every month has been quietly handing you a worse version of itself. It was never broken. It was aimed. Read this: arxiv.org/abs/2406.17737

Hey @SingaporeAirlines! Could someone tell me what the sweet dish was for 12F on flight SQ 967 on May 13th? Thanks in advance! 🙏 SQ967 #SingaporeAirlines #FlightFood

Not really. The Palace on Wheels (with original Maharajas' carriages) on metre gauge was still running into the early 1990s. I travelled on it in 1986 (?) on the steam locomotive from Delhi Cantt to Bandikui.

Tanla is proud to be recognised with Special Recognition at the @iitmadras Social Impact Awards 2026 for the Cyberabad Traffic Pulse initiative, marking the second consecutive recognition for the initiative since its launch. The award was received by Jitendra Kumar Singh and Uma Singh. Developed as a joint initiative with @SCSC_Cyberabad and @CYBTRAFFIC, the platform enables the delivery of verified, real-time traffic and safety alerts including congestion, diversions, accidents, and emergency updates directly to citizens across WhatsApp, RCS, and SMS. In two years, it has delivered 25 million+ messages annually, with 31,000+ subscribers relying on it as a primary source for road safety and civic information. Delivered as a CSR initiative at zero cost to citizens and government, it reflects how technology can scale as public infrastructure, driving safer roads, better commute decisions, and stronger citizen engagement. #CSR #SocialImpact #ESG

@NostalgicHubb Nameds?

Can anyone crack this code? What was the fathers name?

The @urbancompany_UC have replaced their customer service with a bot that cannot comprehend the simplest of issues. I informed it about my intention to sue the company, which it acknowledged but could not offer any solution :-) Gentlemen, think about it and consult your lawyers.

@ashishchauhan Congratulations, Ashish!! 🙂

Ashishkumar Chauhan: The man who built NSE and came back to rebuild the trust. The NSE MD & CEO helped rebuild the exchange and its reputation after a period riddled with scandals and governance lapses. All eyes are now on the IPO forbesindia.com/article/leader…

Anthropic leaked 512,000 lines of Claude Code source code yesterday. What happened in the next 12 hours is absolutely wild. 4 AM. Anthropic pushes an update to npm. Inside the package: their entire codebase. A 60 MB debugging file accidentally bundled in. 23 minutes later, researcher Chaofan Shou spots it. Downloads the zip. Posts it on X. Within 6 hours: 3 million views. By the time Anthropic’s team woke up, the code was forked 41,000+ times across GitHub. Anthropic started firing DMCA takedowns. Too late. A Korean developer named Sigrid Jin woke up to his phone exploding. He’s Claude Code’s biggest power user. WSJ reported he burned through 25 billion tokens last year. He read the leaked code. Rewrote the entire thing in Python in 8 hours. His repo hit 30,000 stars faster than any GitHub project in history. Then he rewrote it again in Rust. That version now has 49,000 stars. Someone mirrored it to a decentralized platform with one message: “will never be taken down.” The code is permanent. Anthropic cannot get it back. Here’s the part I can’t stop thinking about: Anthropic built something called “Undercover Mode.” Its only job: prevent Claude from accidentally leaking internal secrets. They shipped an entire anti-leak system in their own product. Then leaked their own source code in a .map file. Irony is beautiful

@Mic_James_W @PhysInHistory If it was possible to prove that a proof must exist, that by itself would constitute a direct proof of the conjecture. Or would it not?

@PhysInHistory This actually raises an interesting question: is it actually possible to 'prove' whether a mathematical conjecture is possible to prove or not? E.g. Regardless of whether the twin prime conjecture is true or not is it possible to prove that a proof is possible or impossible.

Imagine solving 357 years old mathematical problem that other mathematicians said "impossible to prove".

The Ministry has issued the following directions through the Directorate General of Civil Aviation (DGCA): 1. Minimum 60% of seats on any flight to be allocated free of charge to ensure fair access 2. Passengers travelling on the same PNR to be seated together, preferably in adjacent seats 3. Carriage of sports equipment and musical instruments to be facilitated in a transparent and passenger-friendly manner, subject to applicable safety and operational regulations. Airlines shall also bring out clear, transparent policies for carriage of pets. 4. Strict adherence to passenger rights framework, particularly in cases of delays, cancellations and denied boarding 5. Prominent display of passenger rights across airline websites, mobile applications, booking platforms, and airport counters 6. Clear communication of passenger entitlements in regional languages to ensure wider accessibility and awareness The Ministry of Civil Aviation remains committed to enhancing passenger experience, ensuring transparency, reducing grievances and upholding the highest standards of safety across the aviation ecosystem. @RamMNK @mohol_murlidhar @samirsinha69 @AAI_Official @DGCAIndia @Pib_MoCA pib.gov.in/PressReleseDet…

i watched a flight go from $483 to $547 in 24 hours WITHOUT a single seat selling searched london to new york on a tuesday $483 checked again 2 hours later $512 next morning: $547 panicked and booked it the guy sitting next to me paid $391 same seat, date + airline $156 less he searched once i searched 3 times the algorithm saw me come back and charged me until i broke the seat doesnt have a price you have a price and it goes up EVERY time you show interest couldnt stop thinking about it so i tracked down someone who actually built pricing algorithms for a european carrier asked him what happened to me "you got profiled. the system assigned you an intent score after your second search and raised your ceiling every time you came back" asked how to beat it "most people think a VPN fixes it. thats 2015 advice. the algorithm fingerprints more than your IP now. it reads your device your browser your screen resolution your timezone. VPN to bucharest but your clock says london and your language is english? the algo knows youre faking and sometimes charges you more for trying" "so what actually works?" "you have to poison the entire profile. not just the location. the identity" the protocol he gave me: VPN AND match your timezone and language to the spoofed location. mismatched signals flag you and can trigger a price increase use a fully clean browser. no history no saved passwords no google account. the algorithm fingerprints your session not just your cookies one search one booking. the intent score activates on the second search. there is no safe way to look twice book tuesday or wednesday 1-5am. lowest traffic means the least demand data for the algorithm to inflate against if the price already spiked go dark for 72 hours minimum. not 24. the intent score on most carriers decays on a 3 day cycle. come back on a different device from a different network "we spent $4 billion building these systems. theyre not going to lose to someone who opened an incognito tab" $900 billion industry the gap between what you pay and what the person next to you pays is not a bug its the entire business model stop letting an algorithm charge you for being predictable

My son @shawnchauhan1 start up Sypha.ai was invited for the #CEO roundtable at @PMOIndia on February 20,2025. Shawn was privilged to speak to Hon’ble @PMOIndia Shri @narendramodi ji and explain the work his company Sypha.ai is doing in vibe coding and other areas. #AISummit

The math on this project should mass-humble every AI lab on the planet. 1 cubic millimeter. One-millionth of a human brain. Harvard and Google spent 10 years mapping it. The imaging alone took 326 days. They sliced the tissue into 5,000 wafers each 30 nanometers thick, ran them through a $6 million electron microscope, then needed Google’s ML models to stitch the 3D reconstruction because no human team could process the output. The result: 57,000 cells, 150 million synapses, 230 millimeters of blood vessels, compressed into 1.4 petabytes of raw data. For context, 1.4 petabytes is roughly 1.4 million gigabytes. From a speck smaller than a grain of rice. Now scale that. The full human brain is one million times larger. Mapping the whole thing at this resolution would produce approximately 1.4 zettabytes of data. That’s roughly equal to all the data generated on Earth in a single year. The storage alone would cost an estimated $50 billion and require a 140-acre data center, which would make it the largest on the planet. And they found things textbooks don’t contain. One neuron had over 5,000 connection points. Some axons had coiled themselves into tight whorls for completely unknown reasons. Pairs of cell clusters grew in mirror images of each other. Jeff Lichtman, the Harvard lead, said there’s “a chasm between what we already know and what we need to know.” This is why the next step isn’t a human brain. It’s a mouse hippocampus, 10 cubic millimeters, over the next five years. Because even a mouse brain is 1,000x larger than what they just mapped, and the full mouse connectome is the proof of concept before anyone attempts the human one. We’re building AI systems that loosely mimic neural networks while still unable to fully read the wiring diagram of a single cubic millimeter of the thing we’re trying to imitate. The original is 1.4 petabytes per millionth of its volume. Every AI model on Earth fits in a fraction of that. The brain runs on 20 watts and fits in your skull. The data center required to merely describe one-millionth of it would span 140 acres.

This is not a CBI facility. Nor the office of Mumbai Police. Reuters visuals from an abandoned scam center in Cambodia shows how scamsters used fake signs, Indian flag, and even pictures of Gandhi and Ambedkar to defraud Indians from abroad in "digital arrests".

In a twist of irony, Dr @rssharma3, who was the first Mission Director of UIDAI that provides digital identity (#Aadhaar) to 140 crore people, must now prove his identity during SIR!

@MuditKakkar5 @parkplus_io Service providers often optimize for everything but customer experience, which consequently gets relegated to their ads and marketing copy.

@parkplus_io @sunilbajpai Same is in my case @parkplus_io . Paid for the challan 60 days ago…still not settled, I am given a new date every time I raise a ticket and the ticket is marked as resolved…I am filing a consumer complaint now for this unprofessional conduct

Dear @parkplus_io - You've been repeatedly sending me a messages with "Park+ App → My Orders → Order Details → Verify OTP", but that damn OTP process never competes. This is harassment.

Surprised this wasn’t bigger news in India today. Following their caucus meeting yesterday, the New Zealand Labour Party has officially announced it will support the enabling legislation for the New Zealand-India Free Trade Agreement (FTA). This decision effectively "rescues" the deal for Prime Minister @chrisluxonmp , as his coalition partner, New Zealand First, remains staunchly opposed and has invoked the "agree to disagree" clause.  Labour says it’s a continuation of their long-standing efforts to deepen ties with India and that it would be irresponsible to block a deal with the world's fastest-growing major economy. Though they remain concerned about the "dairy deficit" (the exclusion of core dairy products from the FTA), Labour argued that some progress is better than none and that the deal provides a vital foothold for other sectors like education, tech, and wine. The ruling National Party offered certain "political concessions" to secure Labour's votes. A specific one-year review mechanism is built into the agreement. This allows for a formal pursuit of "further improvements", specifically regarding dairy and environmental standards. The migration provisions, which are important for India, will also get special scrutiny. The enabling legislation will now move through a parliamentary Committee process and is expected to pass with a "super-majority" (National + ACT + Labour) in the first half of 2026. We may be in business with New Zealand by the end of the year! NZ may be a small economy, but concluding an FTA with it will be game-changer symbolically and signal that India is increasing its trade channels — and widening its options. Now it’s time to turn our attention to the first #INDvNZT20I!

India’s future is bright! 🇮🇳 Huge congrats to the winners of 5th School Youth Ideathon 2025 at @IITDelhi. These young innovators wowed us with solutions for a better tomorrow.🏆 Organized by @ThinkStartup, @MEPSCindia, & @cbseindia29 Powered by @TanlaPlatforms The winners are:

@parkplus_io 9810422383. The messages come on WhatsApp.

@sunilbajpai We apologize for inconvenience caused. Could you please share your phone number with us via DM. Alternatively, you can reach out to us at support@myparkplus.com. will assist you further. Thanks for your understanding.

@thebetterindia Chikoo from Gholvad?

It’s soft, it’s juicy, and it’s got just the right touch of sweetness. But what really sets this fruit apart is where it comes from. Tucked away in Maharashtra is a small village that’s become famous for growing a fruit so unique, it’s earned a GI tag. You won’t find massive farms or commercial plantations here. Just generations of local farmers who’ve mastered the art of growing something truly special. From local markets to international kitchens, this little fruit has found fans everywhere. And while it might not look flashy at first glance, one bite is all it takes to understand the hype. Know what we’re talking about? Drop your guess in the comments! #GuessTheFruit #maharashtra #GITagfruit #FruitsofIndia [Purandar village, Maharashtra figs, GI tagged fruit, Indian farming heritage, local produce, global favourite, soft and sweet figs]