Notable verbatim quotes

Each quote is taken verbatim from transcript.md. Some preserve speech-to-text artifacts; bracketed glosses indicate likely intended wording.

On what AI is (one-sentence definition)

"It is simply a new kind of software that is good at pattern recognition." — Section: The one-sentence definition

"people sometimes start thinking that these are magical machines that can think and reason like a human being. And I'm still convinced that they are not." — Section: The one-sentence definition

On why AI took 70 years

"AI is having a birthday in two weeks June 18th. Artificial intelligence is turning 70 years old." — Section: AI's birthday: Dartmouth 1956

"only now did we get enough computing power and enough data to train good AI." — Section: AI's birthday: Dartmouth 1956

"If you if you try to build the eye with insufficient computing power it needs efficient data you do not end up with artificial intelligence you will end up with artificial obnoxiousness." — Section: AI's birthday: Dartmouth 1956

On pattern recognition as the real shift

"for the first time ever we had software that was good and packing up initiation [pattern recognition] it was a new superpower something we never had." — Section: Pattern recognition as a new superpower

On the smartphone vs Apollo 11

"in our pockets we literally carry 100,000 fold of the computing power that we need for a man to move landing. And we use this to slide fruit candy in a row tenant explodes." — Section: AI in your pocket: the smartphone

On AI as a tool that cuts both ways

"It's one of these many examples of technology that give you the choice if it's going to make you dumb or smarter. Choose wisely" — Section: PhotoMath & education

On the human brain as the model

"if 30 years ago you needed a good pattern recognition machine you could easily hire one. We called it an employee." — Section: How brains do pattern recognition

"Our brain is one big sponge of connected neurons. 86 billion neurons in one human brain" — Section: How brains do pattern recognition

On what training actually is

"neurons that get activated together build stronger and stronger connections." — Section: How brains do pattern recognition

On the black box

"Now how does it do this so why does this resulting network apparently react with the word cat to any possible picture of a cat … nobody knows. We have no idea. But it works through even. That's how in Belgium we deal estate politics. It somehow functional don't touch it in the fall apart." — Section: The black box of AI

"before you start writing hardcoded software the human has done the thinking … here we didn't find the solution. We just told the computer adapt this neural network until it works" — Section: The black box of AI

On training-data bias

"that's how they found out that to tell wolves and hospitals [huskies] apart the software was looking at one thing only. Is there snow in the background?" — Section: Training-data bias: wolves vs huskies

"Humans are very bad we have a blind spot for this kind of bias. It is so obvious that we do not see it." — Section: Training-data bias: wolves vs huskies

"the main thing that the computer was looking for was is there a ruler of a doctor next to this part?" — Section: Training-data bias: skin cancer ruler

On hobbyist AI

"Any teenager that wants to walk on their own door that only opens when it hears their voice. Can build this today. It's not that difficult." — Section: Hobbyist AI: Custom Vision & Teachable Machine

On LLMs as imitation, not thought

"I am not sure that AGI will lead to AGI [likely 'that AI will lead to AGI']. I think it is very impressive language imitation. I don't think it does it has the architecture of conceptual thinking" — Section: From pattern recognition to generative to (?) AGI

"when I ask a lot of language model to write me a story about a cat, I'm convinced. Doesn't have any conceptual idea of what a cat is. But the word cat is a data point that statistically follows the previous data point." — Section: Wrap-up: LLMs as language imitation, not thought

"you can ask large language model to write a summit [sonnet]. About an animal and a certain emotion and what comes out absolutely blows my mind. So a lot of room for debate on how much conceptual thinking" — Section: Wrap-up: LLMs as language imitation, not thought