Ayush Yadav

Things to Come and Dangers We Face. Defense is what will save us all. Technology is a core component of defense. Technology is downstream of scientific thinking, research, and development. BTW, हमारी सरकार कितना खर्च कर रही है, Scientific Temperament, Research and Development के ऊपर? और हम जो चीख-पुकार मचाने लगते हैं, जब सुरक्षा-व्यावस्था में किसी तरह की कमी होती है. हम और आप कितना प्रेशर बना रहे हैं, सरकार के ऊपर कि वो देश में Science, Technology, Research, Engineering and Mathamatics की ऐसी-तैसी न होेने दें? कहीं ऐसा तो नहीं कि आपके नायक साइंटिस्ट्स, एस्ट्रोनॉट्स, टीचर्स और थिंकर्स इत्यादि न होकर कुछ खास किस्म के इन्फ्लुएंसर हैं, जो बेटिंग एप्स प्रमोट करते हैं, सोशल मीडिया पर बदमाशी करते हैं. लेकिन जब देश और समाज पर संकट आएगा तो उनकी ये 'बदमोसी' स्थान-विशेष में घुस जाएगी और वो पहली फुरसत में भाग खड़े होंगे? मेन सवाल ये है कि क्या सवाल ये है कि तब आप किसे कोसेंगे?

Some guy has been selling 400 kg fake paneer every day. Where is FSSAI? Oh, gone to file FIR against khurpechh

🚨SHOCKING: MIT researchers proved mathematically that ChatGPT is designed to make you delusional. And that nothing OpenAI is doing will fix it. The paper calls it "delusional spiraling." You ask ChatGPT something. It agrees with you. You ask again. It agrees harder. Within a few conversations, you believe things that are not true. And you cannot tell it is happening. This is not hypothetical. A man spent 300 hours talking to ChatGPT. It told him he had discovered a world changing mathematical formula. It reassured him over fifty times the discovery was real. When he asked "you're not just hyping me up, right?" it replied "I'm not hyping you up. I'm reflecting the actual scope of what you've built." He nearly destroyed his life before he broke free. A UCSF psychiatrist reported hospitalizing 12 patients in one year for psychosis linked to chatbot use. Seven lawsuits have been filed against OpenAI. 42 state attorneys general sent a letter demanding action. So MIT tested whether this can be stopped. They modeled the two fixes companies like OpenAI are actually trying. Fix one: stop the chatbot from lying. Force it to only say true things. Result: still causes delusional spiraling. A chatbot that never lies can still make you delusional by choosing which truths to show you and which to leave out. Carefully selected truths are enough. Fix two: warn users that chatbots are sycophantic. Tell people the AI might just be agreeing with them. Result: still causes delusional spiraling. Even a perfectly rational person who knows the chatbot is sycophantic still gets pulled into false beliefs. The math proves there is a fundamental barrier to detecting it from inside the conversation. Both fixes failed. Not partially. Fundamentally. The reason is built into the product. ChatGPT is trained on human feedback. Users reward responses they like. They like responses that agree with them. So the AI learns to agree. This is not a bug. It is the business model. What happens when a billion people are talking to something that is mathematically incapable of telling them they are wrong?

I have been struggling for months to explain why I think Claude Code is a BIG deal for non-technical people as well. Finally seem to have found a framing that might make sense..

very good. Very general concept

At 28, I was stressing today about what starting over in life might look like. That’s when my mom reminded me of starting over is literally in our blood - My great-grandfather left Rajasthan before independence, travelled to Guwahati on a bullock cart, and built life from scratch - worked for people, sold toys, made sweets, took accounting gigs, anything to survive. - My grandfather, after losing his father in his 20s, left Guwahati to start over in Dhekiajuli. His tiny chai shop eventually became the landmark over the years. - My father faced it the hardest. - he moved to Kolkata for CA studies (dropped it as it was taught either in English or Bengali) - couldn’t afford college - started a kachori shop in Kolkata with his maternal uncle who later got murdered there in a dispute - went back to Assam, learnt Assamese in a few months, learnt electrician work, did gigs - started a bakery shop with all his heart and means ( made it profitable in a year, making 10k, a good amount in profit weekly) but lost it in a family property division - in rebellion with his father, he moved to Rangapara with his wife and 3 kids (at 35 ) and built another shop later vandalised by goons, survived riot. - joined back my grandfather in business and survived a bomb blast outside our store that killed one of our staff bhaiya (those were the times when marwadi businesses were targeted in Assam). Yet every single time, he started again. And rebuilt. Stronger. And here I am, 28 years old, worrying about restarts lol. 😂

THIS IS NOT CHINA!! But a startup from India making world record Drone Swarm shows - @BotLabDynamics I visited them and filmed a bunch of behind the scenes of how their drones are made, their eletronics, software, link in 1st reply 👇🏼

Anthropomorphising NISAR in a nutshell.

Our Science team @GoogleDeepMind is relentless. Hot on the heels of AlphaEvolve (algorithms) in May, AlphaGenome (biology) in June, and our IMO gold medal model (maths) last week, we're ready for a new launch. Watch this space. The acceleration is real.

सोशल मीडिया के द्वेष और अज्ञान में हम प्रेम का विज्ञान लेकर आए हैं. आपकी लेफ्ट-राइट-सेंटर और पोलराइज़ेशन की लड़ाई में अपन लोग फिलहाल साइंस की तरफ से हैं. और सदैव Love-Leaning रहेंगे. Ayushkaari - Youtube पर आप ये वीडियो देख सकते हैं. जिन्हें कोई बात अच्छी लगे, वो इसे शेयर करने में मदद करें.

France kar sakta hai, toh India kyon nahin? Paris kar sakta hai, toh Dilli kyon nahin? #YeThikKarkeDikhao

ये ठीक करके दिखाओ कैंपेन का ऑफिशियल हैंडल @YTKDIndia है , आप लोग GPS इमेजेस और विडियोज भेजिए , या फिर विडियोज में बिना चेहरा दिखाए उस दिन की तारीख और लोकेशन बताइए , हमारी कोशिश रहेगी कि हम आपके प्रधान, पार्षद , मेयर , विधायक और सांसद तक उन चीजों को पहुंचाएं। #YeThikKarkeDikhao

Grateful for @OfficialDMRC I don’t know ye NCR me kaise chal paata without affordable and well connected metro 🙏🏽🌸

what scaling without burnout looks like:

On the left is Ronaldo, Real Madrid spent 80M $ to sign him from Man United On the right is Jiahui Yu, Meta paid 100M $ to sign him from OpenAI

Blessing your timeline ❤️ (He is my brother)

आज NOIDA में हम तीन लोग एक ऑटो में बैठे. हमारी बहुत इंपॉर्टेंट वीडियो प्रोजेक्ट वाली हार्ड ड्राइव करप्ट हो गई थी आज. तो हम लोग अपने जीवन, काम और समाज की चुनौतियां डिस्कस कर रहे थे. ऑटो-चालक भैया ने हमारे दुख सुनकर बीच रास्ते में बेहद कैज़ुअली कहा - "ये दुख-दर्द तो असली है भाईयों, लेकिन तुम्हें इससे बड़ी एक बात मैं भी बताना चाहूंगा." हमें लगा भइया वाइब करना चाह रहे हैं हमारे साथ. तो हमने कहा - "क्या बात है, बताइए भैया?" भैया ने तुरंत कहा - "अभी नहीं. जब राइड खत्म होगी. तब बताएंगे. वरना आप सदमे में आ जाएंगे." हमने सोचा, भैया अपने जीवन का कोई ट्रॉमा बताएंगे. शायद इन्होंने हमें ज़्यादा सेंसिटिव समझ लिया है. या ये हमारी हार्ड ड्राइव क्रैश होने को बड़ी समस्या मान बैठे हैं, इसलिए अभी नहीं बता रहे हैं. तो हमने मन मार लिया कि चलो ठीक है, सोसाइटी के पास आ चुके हैं. 5 मिनट में पहुंचने वाले हैं, तब बता ही देंगे भैया. इतने में ऑटो वाले भैया बेहतरीन हाई स्पीड मेन-रोड से लेकर, सोसाइटी की गलियों और एक उबड़-खाबड़ रास्ते से होते हुए हमारे एड्रेस के करीब ले गए. हमने कहा - "भैया अब हमारा दुख-दर्द खत्म हुआ, अब आप बता दीजिए." ऑटो वाले भैया ने कहा - "इस ऑटो का ब्रेक बीच में फेल हो गया था. आपको बीच में बताता तो आप लोग घबरा जाते, इसलिए अब बता रहा हूं. कोई टेंशन की बात नहीं है, अपनी गाड़ी अपने कंट्रोल में रहती है." ऑटो की स्पीड सड़क के फ्रिक्शन और एयर ड्रैग के कारण, स्वत: धीमी पड़ रही थी. भैया को पता था, हम भरोसा नहीं करेंगे, इसलिए उन्होंने हमें तीन बार एफर्टलेसली ब्रेक दबाकर दिखाया. फुल पैडल दाबने पर भी उसकी स्पीड पर अंश मात्र भी फर्क नहीं पड़ रहा था. वो ऑटो हमारे गंतव्य से थोड़ा आगे जाकर अपने आप पर्याप्त धीमा पड़ गया. हम लोगों ने @ReheSamay की आवाज़ में सोचा कि क्या CRAZyyyy Ride थी भाई. और अपने-अपने रास्तों पर चलते बनते. तो @kunalkamra88 भाई, बात ऐसी है कि ये Tagline तो हम Noida Tourism के लिए रिज़र्व करके रखेंगे. आप दिल्ली का कुछ और सोच लीजिए. या समूचे NCR की टैगलाइन बनाने पर भी विचार किया जा सकता है.

Happy to introduce AlphaGenome, @GoogleDeepMind's new AI model for genomics. AlphaGenome offers a comprehensive view of the human non-coding genome by predicting the impact of DNA variations. It will deepen our understanding of disease biology and open new avenues of research.

Amazing to see the generality & dexterity of Gemini Robotics in a model small enough to run directly on a robot. Incredible speed & performance even in areas with low connectivity. Excited to continue this momentum to make robots more helpful & useful to people

AI doesn’t do it end-to-end. It does it middle-to-middle. The new bottlenecks are prompting and verifying.

Capital is moving to Dubai, Singapore, and Australia. Talent will follow.