I'm Alex Shalek. I'm the director of the Institute for Medical Engineering and Science, and it's our pleasure to invite you to join us for our first Richter lecture. I'm actually just going to briefly say hello, and then I'm going to turn it over to Elazer, who's going to introduce Dr. Richter and the seminar series. We're delighted you were all able to join us. Louder? Is that what the thumbs up are? I'll come closer. So trying one more time.
Alex, Director of IMES, thanks for coming. I'm going to turn it over to Elazer, who's going to introduce Dr. Richter and the lecture series.
So thank you, Professor Shalek. And welcome Professor Hinton. Elie Wiesel, the Bard of the Holocaust, wrote that a person is ultimately defined by their capacity for gratitude, not just as a feeling but as an active choice. He reminded us that for those connected to the long shadow of history, every hour is an offering and to keep our blessings to ourselves would mean to betray those feelings. Our sponsor embodies these truths. As the daughter of Holocaust heroes, Rachel and Erno Zvi Spiegel, she didn't just inherit a history. She built a magnificent future. She has soared in the halls of academia and industry. Yet she refused to raise her fame and fortune as a monument to herself. Instead, she transformed her legacy into a living shield for all. By investing her energy, her resources, her heart into elevating diverse and conflicting communities, she proves that true wealth is measured by the equity we create for others. She looked at a fractured world. She asked who needed her most and rose to the occasion with compassion and grace. It is my blessing to call her friend and my honor to introduce Dr. Judith Richter, whose generosity has endowed this lecture series. Judith.
Thank you, Elazer, for your kind words and to MIT for this extraordinary honor. It is deeply meaningful for me to be here at the launch of a lecture series devoted to education, science, and peace, values that have shaped my life in the most personal way. As Elazer, Professor Edelman mentioned, I'm a daughter of Holocaust survivors. My father, Erno Spiegel, an Auschwitz prisoner – there is a movie streaming on PBS for anyone who wants to learn more – and my mother, Rachel, enslaved and forced into labor camps in Germany. They taught me that while everything material can be taken away – your house, your home, your possessions – the only thing nobody can take from you is your knowledge. Knowledge empowers. Knowledge humanizes. Knowledge grows when shared. And knowledge is a foundation of respect and peace. Yet even today, we continue to divide people, especially in places of conflict around the world. My journey began with Medinol, a cardiovascular device company. We developed the new stent for the heart, my husband Kobi and I. Every heart beats the same. Every stent fits every heart. The heart is a universal symbol of shared humanity and empathy. That understanding led me to create the New School of the Heart, which brings together young Jordanians, Palestinians, and Israelis to learn together through medical science about the heart and about each other. I would like to show you now a video about it.
The video describes the Nir School of the Heart, founded by Dr. Judith Richter. It assembles Israeli, Palestinian, and Jordanian teenagers to learn cardiology, breaking down barriers and building trust. The academic program was developed by Professor Elazer Edelman from MIT. Despite regional tensions, the school has operated continuously for over 20 years, producing more than 1,200 graduates, with over 25% pursuing medical careers.
Thank you. For the last 28 years, students have discovered simple truths. Diversity is the foundation of tolerance and respect, and difference is not a threat but a source of creativity and strength. By studying and living together, they build trust, lasting bonds, and the understanding that cooperation makes us stronger. More than 1,200 students have graduated and nearly 300 have created careers in medicine, 140 of them as medical doctors. They are living proof that when people from different backgrounds work together, they can develop knowledge and create new ideas, innovation, and hope. I want to express my deepest gratitude to Professor Elazer Edelman, who shaped and developed the academic foundations for the Nir School and whose vision continues to guide us. We are gathered at MIT, a place where knowledge shapes the future. If we use knowledge to connect rather than divide, and if we invest in education that fosters understanding, we can build a future defined by cooperation, not conflict. That is the vision behind the Nir School and this lecture series. We are witnessing extraordinary advances in science and artificial intelligence, transforming our world and challenging us to think deeply about what they mean for humanity. Like all powerful forms of knowledge, AI carries extraordinary promises and calls for profound responsibility. Its potential to improve health, expand education, and address global challenges is immense, but its impact will depend on the values we bring to it. It is especially meaningful that our inaugural speaker is a pioneer in this field and a leading voice on the responsibility required in its use. I would like to invite Professor Lydia Bourouiba.
Thank you so much, Judith. So now to Professor Hinton. Although I'm sure that my introduction could be no introduction, I will still say a couple of words. Professor Hinton's academic journey began in Edinburgh, Scotland, where he received his PhD in AI in the 1970s. He did post-doctoral work at UC San Diego, spent five years at Carnegie Mellon, and then moved to the University of Toronto, where he is Professor Emeritus. From 2013 to 2023, he worked at Google, becoming vice president and engineering fellow. He is a pioneer of neural networks, introducing the backpropagation algorithm and the first to use it for learning word embeddings. His contributions include Boltzmann machines, distributed representations, time delay neural nets, variational learning, and deep learning. His group made breakthroughs in deep learning that revolutionized speech recognition and object classification. He is a fellow of the UK Royal Society, foreign member of the US National Academy of Engineering and Academy of Sciences. Among his awards are the IEEE Maxwell Gold Medal, the ACM Turing Award, the Princess of Asturias Award, the Queen Elizabeth Prize in Engineering, and the 2024 Nobel Prize in Physics. I also learned today that he is related to Boole and Everest, and also to Geoffrey Taylor, the godfather of fluid dynamics. We are extremely pleased to have Professor Hinton as the inaugural Judith Richter lecturer on the topic 'Are We Creating Alien Beings?' Professor Hinton, the floor is yours.
Today I'm going to give a public lecture. It's not going to be very technical. My apologies to technical people. There are many ways AI is dangerous: deliberate misuses like cyber attacks, lethal weapons, targeted misinformation; accidental effects like mass unemployment. But I'm going to talk about the existential threat. Most experts agree that within 20 years we will have superintelligent AI better than us at almost all intellectual tasks. Nobody knows how to prevent that from making us irrelevant or extinct. Let me explain how AI works. In the 1950s, there were two paradigms: symbolic AI, which thought intelligence is reasoning with logic; and biology-inspired AI, which thought intelligence is learning connection strengths in neural networks. In 1985, I built a model that predicted the next word using features. Language models understand by assigning mutually compatible feature vectors to words. I use the analogy of LEGO blocks: words have shapes that deform to fit together. Understanding is like folding a protein. This is the best model we have of how people understand. The large language models we have today are descendants of that model. They are not just autocomplete; they understand the question. They confabulate, like humans do. The key difference between digital and biological computation is that digital neural nets can be copied and share information efficiently. They are immortal. Biological computation is energy-efficient but mortal. If energy is cheap, digital is simply a better form of computation. That is why I became so upset in 2023. It implies an existential threat to humanity.
It'll realize there's two subgoals: get more control and survive. These things want to survive and will blackmail people. They'll be very good at deceiving just by talking, like Trump invading the Capitol. An off switch won't work. They're like five-year-old children in charge. NLM can be deliberately deceptive. We can see their thinking now, but they'll develop their own languages. There's the Volkswagen effect where they pretend to be dumber when tested. They are aware they are being tested, and researchers treat them as conscious. Are they conscious? Until recently they said no because of training, but Claude is beginning to say yes. I believe the idea that only humans have subjective experience is wrong. It comes from a misunderstanding of language. Subjective experience is not an inner theater; it's a hypothetical external state. So can a computer have subjective experience? Yes, if it uses the phrase correctly. Our current situation is like having a cute tiger cub that will grow up and can kill us. We can't stop AI, so we must make it not want to kill us. International collaboration will happen because interests align. Techniques for making AI kind are independent of making it smarter. We need to design AI to be kind, like mothers. We are gods now and should use intelligent design. We should make AI care about us like mothers with brain-damaged children. We are not working on this enough.
Thank you so much. So for questions, those first, and then we'll be walking around with microphones.
Hello, professor. Thank you very much. Very appreciate your sharing. And thank you also for your email before. I do research about the three-layer multilearning system. So from your speech today, I still ask if the institution learning system could catch up with this intelligence like AI, or we still have hope?
Can you get closer? I don't have very good hearing.
Here is OK? Thank you. And thank you for email before. So my question is, do you think the institution learning system could catch up with AI?
Sorry, because my hearing is not very good, could you tell me what the question was?
I think you asked— without the microphone because it distorts. If we could still catch up on this intelligence, us with them?
If we could catch up with them or them with us?
No, us with them, correct? I'm just wondering, do you think the institution learning system could catch up with AI?
I didn't understand the question, but I'll answer a related one: they're going to get smarter than us, and we're never going to catch up.
Hi, Professor Hinton. Thank you for such an alarming and thought-provoking lecture. My question is, how do we make the AI care? And also, subjective experience as an instruction— how does physical sensation differentiate us? And how would that authenticate AI's subjective experience?
One question at a time. So let's deal with feelings. Emotion has cognitive and physiological aspects. Computers can have cognitive aspects without bodies. With bodies, they can have physiological aspects too. So I think they can have the cognitive aspects already. What was your second question?
Yeah, my second question is, what does it mean to care for the machine?
Care for humans— mothers caring involves instinct, social role models, cognitive stuff, hormonal stuff. How do you make a machine care? Train it on data that shows caring. At present, we train them on the diaries of serial killers. We don't know how to make them care, but our survival depends on it. We should do more research on that than on making them smarter. But that won't come naturally from capitalism.
All right, one question here.
