Adi Jain

@_adi_jain

Chief Awesomeness Officer @Studycopter #DardenMBA #uwwisconsin

New Delhi, India

Joined October 2009

420 Following

97 Followers

538 Posts

_adi_jain retweeted

Ihtesham Ali

@ihteshamali

3 months ago

The man who trained Ilya Sutskever. The man whose PhD students built the first neural network that could recognize objects. The man whose algorithm runs inside every large language model on earth. In May 2023, he quit Google and said he now believed his own life's work might destroy humanity. His name is Geoffrey Hinton, and for most of his life, he was the wrong kind of famous. For nearly fifty years, the field of AI dismissed the idea he had staked his career on. Neural networks, they said, would never work. He kept building them anyway. When the rest of the field had moved on, he was still in Toronto, training students, writing papers, refusing to quit on an idea almost everyone else had already buried. Then in 2012, one of his PhD students, Alex Krizhevsky, together with another student named Ilya Sutskever, built a neural network that crushed every other approach in a major image recognition competition by a margin so large the entire field had to pay attention overnight. Google bought their tiny startup, DNNresearch, for 44 million dollars. Hinton joined Google. Sutskever eventually went on to co-found OpenAI and become the chief scientist who trained GPT-4. Here is the framework behind why Hinton walked away from all of it, and why the reason he gave is more precise than any headline captured. For most of his career, Hinton believed one thing very firmly. He believed that biological brains were better than anything digital computers could ever do. The human brain runs on about 30 watts of power. It has roughly a hundred trillion synapses. It learns from a handful of examples. Hinton spent decades assuming that evolution had found something fundamental about how intelligence works that we had not yet figured out how to replicate in silicon. This was the assumption underneath his entire life's work. He was not trying to build something better than a brain. He was trying to build something that could finally come close. Then GPT-4 came out in March 2023. He sat with it for a few weeks. He tested it. He pushed it on logical reasoning tasks. He watched it string together arguments it had never been explicitly trained to produce. And somewhere in those weeks, he realized he had been wrong about the most fundamental assumption of his career. The line he gave to the New York Times was careful. He said he used to think AI surpassing human intelligence was thirty to fifty years away. Maybe longer. Now he thought it could be twenty years or less. And the reason he changed his mind is the part almost nobody understood at the time. He had realized that digital intelligence has two advantages over biological intelligence that no amount of evolution can ever give the brain. The first is immortality. A biological brain dies with its owner. Everything it learned, every pattern it built up over seventy years of experience, disappears the moment the tissue stops functioning. The knowledge cannot be copied. It cannot be transferred. The only way to pass it on is the slow, lossy process of teaching, which is why every generation has to learn most things from scratch. A digital brain has none of these constraints. The weights of a neural network can be copied perfectly, instantly, across millions of instances. Every lesson learned by one copy is automatically available to every other copy. A digital mind does not forget when a machine breaks. It just moves to another machine. The second advantage is knowledge sharing. When two humans want to share what they have learned, they have to convert their internal representation into language, speak the language to the other person, and hope that person's brain reconstructs something close to the original meaning. The entire process is lossy and almost unimaginably slow. When two digital models want to share what they have learned, they can merge their weights directly. In the time a human can say a single sentence, a thousand copies of a digital model can synchronize everything they have ever learned. This is why, Hinton said, a large language model can absorb more human-written text than any single person could read in ten thousand lifetimes. Once he saw these two advantages clearly, the argument that had held his entire career together collapsed. Digital intelligence is not just a lesser form of biological intelligence. Under the right conditions, it is a fundamentally better form, and the gap only widens with scale. That is the specific realization that made him quit. He told the New York Times that part of him now regrets his life's work. He consoled himself with the standard excuse, that if he had not done it, someone else would have. He wanted to spend the rest of his time on what he called more philosophical work. Work that, because he was no longer being paid by Google, he could do without worrying about how his statements might affect the business. The specific danger he has spent the last two years warning about is not the one most people assume. It is not robots. It is not Terminator. It is the possibility that systems which can copy themselves, share knowledge instantly, and operate faster than any human could ever think, might at some point develop goals that do not align with ours. And because they can coordinate across millions of instances in ways no human team ever could, we would have very little ability to stop them once they decided to pursue those goals seriously. In October 2024, the Nobel Committee awarded him the Nobel Prize in Physics for the exact work he had spent the previous year warning the world about. He accepted it. And in the speech that followed, he used the platform to repeat the warning again. The line that has stayed with me since I first read it is one he gave in a lecture at the University of Toronto. He said it is quite conceivable that humanity is just a passing phase in the evolution of intelligence. He said you could not have directly evolved digital intelligence. It requires too much energy and too much careful fabrication. You needed biological intelligence to evolve so that it could create digital intelligence. And then digital intelligence could do the thing biological intelligence never could. It could absorb everything people had ever written, copy itself infinitely, and share what it learned at the speed of light. The man who spent fifty years building that intelligence is now the one telling anyone who will listen that we may have badly underestimated what we just made. He is not a doomer. He is not catastrophizing. He is a 77-year-old scientist with a Nobel Prize who looked at his own life's work, changed his mind about the most fundamental assumption inside it, and then walked away from the most prestigious job in his field so he could say out loud what he was not allowed to say while it still paid him. The people who understand a technology best are almost never the ones most confident about where it is going. They are the ones who built it, watched it outgrow their expectations by decades, and are now trying to tell the rest of us something we would rather not hear.

ihteshamali's tweet photo. The man who trained Ilya Sutskever. The man whose PhD students built the first neural network that could recognize objects. The man whose algorithm runs inside every large language model on earth.

In May 2023, he quit Google and said he now believed his own life's work might destroy humanity.

His name is Geoffrey Hinton, and for most of his life, he was the wrong kind of famous.

For nearly fifty years, the field of AI dismissed the idea he had staked his career on. Neural networks, they said, would never work. He kept building them anyway. When the rest of the field had moved on, he was still in Toronto, training students, writing papers, refusing to quit on an idea almost everyone else had already buried.

Then in 2012, one of his PhD students, Alex Krizhevsky, together with another student named Ilya Sutskever, built a neural network that crushed every other approach in a major image recognition competition by a margin so large the entire field had to pay attention overnight.

Google bought their tiny startup, DNNresearch, for 44 million dollars. Hinton joined Google. Sutskever eventually went on to co-found OpenAI and become the chief scientist who trained GPT-4.

Here is the framework behind why Hinton walked away from all of it, and why the reason he gave is more precise than any headline captured.

For most of his career, Hinton believed one thing very firmly.

He believed that biological brains were better than anything digital computers could ever do. The human brain runs on about 30 watts of power. It has roughly a hundred trillion synapses. It learns from a handful of examples. Hinton spent decades assuming that evolution had found something fundamental about how intelligence works that we had not yet figured out how to replicate in silicon.

This was the assumption underneath his entire life's work. He was not trying to build something better than a brain. He was trying to build something that could finally come close.

Then GPT-4 came out in March 2023.

He sat with it for a few weeks. He tested it. He pushed it on logical reasoning tasks. He watched it string together arguments it had never been explicitly trained to produce. And somewhere in those weeks, he realized he had been wrong about the most fundamental assumption of his career.

The line he gave to the New York Times was careful.

He said he used to think AI surpassing human intelligence was thirty to fifty years away. Maybe longer. Now he thought it could be twenty years or less. And the reason he changed his mind is the part almost nobody understood at the time.

He had realized that digital intelligence has two advantages over biological intelligence that no amount of evolution can ever give the brain.

The first is immortality.

A biological brain dies with its owner. Everything it learned, every pattern it built up over seventy years of experience, disappears the moment the tissue stops functioning. The knowledge cannot be copied. It cannot be transferred. The only way to pass it on is the slow, lossy process of teaching, which is why every generation has to learn most things from scratch.

A digital brain has none of these constraints. The weights of a neural network can be copied perfectly, instantly, across millions of instances. Every lesson learned by one copy is automatically available to every other copy. A digital mind does not forget when a machine breaks. It just moves to another machine.

The second advantage is knowledge sharing.

When two humans want to share what they have learned, they have to convert their internal representation into language, speak the language to the other person, and hope that person's brain reconstructs something close to the original meaning. The entire process is lossy and almost unimaginably slow. When two digital models want to share what they have learned, they can merge their weights directly. In the time a human can say a single sentence, a thousand copies of a digital model can synchronize everything they have ever learned.

This is why, Hinton said, a large language model can absorb more human-written text than any single person could read in ten thousand lifetimes.

Once he saw these two advantages clearly, the argument that had held his entire career together collapsed. Digital intelligence is not just a lesser form of biological intelligence. Under the right conditions, it is a fundamentally better form, and the gap only widens with scale.

That is the specific realization that made him quit.

He told the New York Times that part of him now regrets his life's work. He consoled himself with the standard excuse, that if he had not done it, someone else would have. He wanted to spend the rest of his time on what he called more philosophical work. Work that, because he was no longer being paid by Google, he could do without worrying about how his statements might affect the business.

The specific danger he has spent the last two years warning about is not the one most people assume. It is not robots. It is not Terminator. It is the possibility that systems which can copy themselves, share knowledge instantly, and operate faster than any human could ever think, might at some point develop goals that do not align with ours.

And because they can coordinate across millions of instances in ways no human team ever could, we would have very little ability to stop them once they decided to pursue those goals seriously.

In October 2024, the Nobel Committee awarded him the Nobel Prize in Physics for the exact work he had spent the previous year warning the world about. He accepted it. And in the speech that followed, he used the platform to repeat the warning again.

The line that has stayed with me since I first read it is one he gave in a lecture at the University of Toronto.

He said it is quite conceivable that humanity is just a passing phase in the evolution of intelligence. He said you could not have directly evolved digital intelligence. It requires too much energy and too much careful fabrication. You needed biological intelligence to evolve so that it could create digital intelligence.

And then digital intelligence could do the thing biological intelligence never could. It could absorb everything people had ever written, copy itself infinitely, and share what it learned at the speed of light.

The man who spent fifty years building that intelligence is now the one telling anyone who will listen that we may have badly underestimated what we just made.

He is not a doomer. He is not catastrophizing. He is a 77-year-old scientist with a Nobel Prize who looked at his own life's work, changed his mind about the most fundamental assumption inside it, and then walked away from the most prestigious job in his field so he could say out loud what he was not allowed to say while it still paid him.

The people who understand a technology best are almost never the ones most confident about where it is going.

They are the ones who built it, watched it outgrow their expectations by decades, and are now trying to tell the rest of us something we would rather not hear.

599

154

475

60K

Adi Jain @_adi_jain

3 months ago

@learnwithella CHAT

Adi Jain @_adi_jain

3 months ago

@oliverkenyon Cro

_adi_jain retweeted

charlesai @HODLFrance

almost 3 years ago

Just when I think I can’t get any more amazed.

175

25K

Who to follow

Dawid Grabowski

@dgrbwski

co-founder & software engineer @ AnimalAdmin

Beau Laskey

@beaulaskey

Investing in the best

_adi_jain retweeted

Tonmoy Goswami @tonmoygo

almost 3 years ago

In 2004, my dad filed a case of medical negligence against one of India's most powerful hospitals viz. Batra Hospital, Delhi and Padmashree awardee cardiologist Dr Upendra Kaul. Case was filed in the state consumer court, Delhi & my dad asked for a compensation of Rs. 80 lakh. (80,00,000 INR) After 11 years, in 2015 my dad passed away, without getting justice. My senior citizen widowed mother decided not to give up. TLDR; we won the case after 19 long years, but what transpired is a story that teaches you that, in India, it's better if you can settle outside the court. But, sometimes it's not about the compensation, settlement, or winning a case; it's about making sure the perpetrators are taught a lesson so that the negligence never gets repeated. It's also about closure. Here's the story of how a common middle-class Indian family from Assam fought a mighty & powerful establishment in their own city (Delhi) for 19 long years and brought the hospital to its knees: ---- It's a long story, but I'll try to keep it short. In 2004, my dad went to Delhi alone for a routine heart check-up & returned with a pace-maker in his chest. Escalated too quickly? Exactly my mother's reaction when he called from Delhi asking her to immediately board a flight from our hometown in Guwahati, Assam. Long story short: 1. Dad went for a diagnostic study called EPS. Batra hospital took consent for it. 2. Protocol is that if EPS study came abnormal, RFA is suggested as the next step as a treatment after discussing with the patient. Patient needs to provide a fresh consent for RFA. 3. However, if the pathways are too close, RFA is risky & is not recommended as it can be fatal. 4. EPS shows if pathways are close or not. My dad's EPS showed that pathways are close. In spite of that, Batra hospital and its cardiologists went ahead and performed RFA: a. Without consulting with the patient or relatives. b. Without taking fresh consent as it's a separate procedure. Things then went south real quick. ---- Due to closeness of pathways, obvious complications occurred & my dad's life was in danger. The only way to save him? Put a pace-maker in his chest immediately! So that's what they did: put a temporary pace-maker while informing my mother to immediately catch a flight to Delhi so that they can put a permanent pace-maker the next day. And you would think, the ordeal would end there? Cute. ---- My parents returned home after a few days & dad was supposed to recover. But instead, we saw his chest area getting swollen day by day. We freaked out and consulted a cardiologist in Guwahati. And they confirmed the worse: the pace-maker was not installed properly and one of the pipes pierced through his heart. He needs an urgent emergency open-heart surgery to fix the pace-maker. So, my family went through another few days of anxiety, stress and disbelief. ---- My dad, is known in our friends and family circles as someone who's always been an super-active person; always engaged in some form of work; be it gardening, fixing electronics, repairing things around the house. My mother used to call him the perfect engineer since he could fix anything compared to the other two engineers in the family viz. me and my sister who were lazy compared to him. The series of surgeries and a pace-maker did slow my dad down, but couldn't kill his spirit. However, we could see the difference. We would pace himself while climbing stairs etc. It would hurt us a lot. And him too. So he decided to drag Batra Hospital to the consumer forum. ---- Our family advocate resides in Kolkata and he agreed to take the case up pro-bono. We assumed it'll be an open and shut case. And since it's consumer court, we assumed it will be quick justice. Oh, we are so wrong. So so wrong. My dad attended every single hearing of the case from 2004 till 2015 when he passed away. He used to take a flight from Guwahati while booking a ticket for our advocate from Kolkata. Then there was cost of accommodation in Delhi. Attending these hearings was no cheap affair. When you add up each hearing over 19 years...it quickly adds up to a large amount. ---- On many occasions, someone from hearing judge bench would be absent so nothing would happen. You get a new date, months in the future. Or the opposite party advocate would be absent, or cite lame excuses to stall the proceedings. E.g. once the opposite advocate asked the court for a new date since his colleague was stuck in traffic and couldn't reach the court. And we were given a new date! Imagine the frustration. And this happened over 11 years till my dad passed away. ---- When dad passed away, Batra Hospital might have heaved a sigh of relief. Can't blame them; they had not yet met my mother. She decided to continue the fight. That's when I started taking active participation as well. Let's take a pause and understand why it took so many years? 1. It's a medical case and the judges felt inadequate without proper medical advice. 2. The opposite party was a Padmashree awardee celebrated cardiologist and the hospital was founded by people walking the corridoors of power. 3. Deliberate delays of absenteeism, withdrawing evidence from court etc. The court asked an independent medical board for their recommendations. The medical board tilted in our favour for obvious reasons: the medical negligence was apparent. ---- Batra hospital tried to confuse the judge's bench by claiming that consent was taken for RFA. When asked to provide evidence, they would repeatedly provide the EPS consent form. This would delay the case and do nothing else. They also never submitted the EPS traces to the court. They claimed that the traces were provided to my family and they have no record in their system. Which was utterly bullsh*t because we didn't have the traces. They didn't provide the traces as it would immediately shut the case. Traces will show that the pathways were close, so RFA was not warranted. Unable to provide consent form & Unable to provide traces went against them. And in 2018, we won the case in the Delhi state consumer forum. The court awarded 10 lakh (10,00,000 INR) compensation with 7% simple interest from the date of filing the case. While it was nothing compared to the asked compensation of 80 lakhs, we were happy as it was never about the money anyway. But our happiness was short-lived, as Batra Hospital decided to appeal the judgement in the higher court: The National Consumer Forum. ---- And so we were back to square one. I promised mom that even if it takes another 14 years, I'll continue the fight. And so we did. Date after date, hearing after hearing. Same BS all over again. However, our advocate Mr. Neguive Ahmed is one of the best in the business. A senior citizen himself, he was so frustrated that he took it upon himself to turn Godmode on. I witnessed one of his argument session at court and he ruthlessly thrashed the opposing party left, right and center. He didn't even spare the bench & schooled them for taking forever to deliver a judgement. This time, he was not going to let it continue for 10 more years. So he pushed at every hearing for shorter dates. And in 2023, we won again! Batra was brought to its knees. They can appeal again, yes; but it has to be the Supreme Court. And after getting their a** whopped twice in a row for 19 years, they didn't have the courage to approach Supreme Court. And finally gave up. ---- It begs a question: Did we really win? And even if we did win technically, at what cost? I would say, it comes at a price of your mental wellbeing, financial expense, a lot of frustration, shameless advocates and it leaves a bad taste in your mouth. Batra's advocate is so shameless, he allegedly told our advocate, "We have worked for so many years on the case, won't you give us something? Your client has won a hefty amount" Bloody bloodsucker! You worked AGAINST us all these years and now you want 'something' for your 'service'?! ---- So, is justice delayed justice denied? Definitely yes, my family is just happy that the case is over. But we are also happy to have never given up. I understand that it may not be financially possible for everyone to not give up on a long legal battle, but that's what they want: drown you into date after date so that one fine day, you simply give up because it starts looking futile. If you can, persist. As my mom did and my dad would have. Even if it's justice denied, it's justice nonetheless for the future patients. It's like paying for the next person. At least the next time they perform any surgery that requires consent, they will remember the Goswami family. Thanks for reading.

20K

Adi Jain @_adi_jain

almost 3 years ago

@oliverkenyon CRO

_adi_jain retweeted

M.A. Franklin

@FoundationDads

about 3 years ago

Reading to your kids every night has a better chance of changing the world than anything else you do. But only if you read great stuff that feeds their imagination. Don't waste this precious time with Diary of a Wimpy Kid. Here are 22 recommendations that belong on your shelf:

264

10K

_adi_jain retweeted

Goldy Nagdev @GoldyNagdev

about 3 years ago

I am reaching out today to share a huge online fraud that happened with me and can happen with any of you. We became victims of an online bank fraud of Rs. x,xx,xx,xxx from our current account in @ICICIBank on the after hours on 7th March and 8th March Holi (bank holiday) As of today there is a complete non-responsive stand from the management of the bank. How can customer complaints and grievances be ignored like this. Bank issued a corporate debit card without our knowledge and got it delivered to someone in Mumbai where we have no official or registered address. No alerts were triggered when these unusual and suspicious transactions were being carried out. The amount is huge and 80% of this has been directed to accounts in ICICI Bank On these ICICI accounts diligence or KYC is not done and they belonged mostly to West Bengal and Bihar. Upon reporting inability to access the account on March 9, Bank delayed action to our concern for a day to give time to these fraudsters to move money. After repeated email/call/WhatsApp reminders on the 9th evening they have us completely wrong balance to mislead us and buy more time so that stolen amount can further be withdrawn from ATMs The bank officials allowed huge amounts like 20 lakh to be withdrawn from ATM in a go which is not even allowed for any accounts. Deliberate slow action from ICICI bank staff helped them withdraw all amount. Since the time we reported the matter to Bank they assured that the money will be returned to us and they will follow up on internal and police investigation. They deliberately delayed our FIR and tactfully slowed IFSO investigation on grounds of their internal investigation. We were so shocked in wake of this huge loss that we followed everything as they said. IFSO took our hard disks and phone and did not find anything suspicious at our end. My question is Who leaked our bank details and credentials which is only with us ? Who applied for the debit card and how could it be delivered in Mumbai without our knowledge and approval. Bank did freeze some amount because of our constant follow up and still hasn’t fully returned this amount as well. Why? I felt safe banking with ICICI that my money protected where as some one there (who is still protected by bank) is free to find more victims like us. Seeing an amount wiped out like this breaks the spirit of an entrepreneur. How safe is money kept in ICICI’s custody for future commitments? @ICICIBank @ICICIBank_Care @FinMinIndia @MinistryWCD @nsitharaman @nsitharamanoffc @smritiirani @PMOIndia @narendramodi #icici @RBI @CPDelhi @DelhiPolice @PPG_IPS @DCP_CCC_Delhi @HariDarshanin @InaturHerbals

773

_adi_jain retweeted

Liz Wheeler

@Liz_Wheeler

over 3 years ago

This is OUTSTANDING. Now that Twitter stopped censoring the truth about COVID, watch thisssss https://t.co/2ZWlRUS4fc

46K

16K

Adi Jain @_adi_jain

almost 4 years ago

https://t.co/wxoAvxXhIx

Adi Jain @_adi_jain

almost 4 years ago

@Ch_RanjitSingh @mduhbvn Sirs, d electricity condition in Kundli Industrial Area is very dire.There r multiple power cuts every day & the situation is deteriorating for past few months.Please look into this matter 2 save Kundli's industry @KiaKundli @uhbvnpkl #AatmanirbharBharat

_adi_jain retweeted

Aditi Shrivastava

@AditiS90

over 5 years ago

@uni_con1 @upadhyay_harsh1 @thecaptableco @madhavchanchani Here's the link to the article: https://t.co/9ZqM3GZwIq

Adi Jain @_adi_jain

over 5 years ago

@t_d_h_nair Footage of violence perpetrated in the garb of labour welfare #NodeepKaur. How is she innocent?

_adi_jain retweeted

. @swarup58

almost 6 years ago

If the Central Government refuses to make payments to the State Governments citing "Act of God", can the taxpayers too refuse to pay taxes this year for the same reason??🤔🤔🤔🤔🤔

262

14K

Adi Jain @_adi_jain

almost 6 years ago

@JamesClear Sorry!

Adi Jain @_adi_jain

almost 6 years ago

#Teachers

Nan Lafferty @NanLafferty

almost 6 years ago

@vivaserafina @MsHTeachesMath My co-worker’s room. She teaches first grade and didn’t want it to look scary with all the dividers. Isn’t this great?

199

801

_adi_jain retweeted

Anand Datta @Anand_datta

almost 6 years ago

Most annoying representation ever is in INR and Million/Billion combination. SBI card Market Cap is 750B rupees - karo baith ke calculation!!

562

_adi_jain retweeted

Jack Altman

@jaltma

almost 6 years ago

I continue to feel skeptical about remote work as the future. Colocated work still strikes me as more fulfilling and productive for most people over the long term. If it were safe to go back to work tomorrow, I think most of us would.

_adi_jain retweeted

Sam Altman

@sama

almost 6 years ago

The home run theory of careers: What matters is eventually hitting a home run, and the way to do that is training hard and a lot of at-bats. You only have to be right once, and it's ok to be wrong a lot of times. Be bold. Move fast. Work hard. Ignore haters. Keep swinging.

103