Federico Faggin: What Distinguishes Us from Artificial Intelligence? | Stesi’s 30th Anniversary

Fine. First of all, thank you for being here, for inviting me to this incredible setting. I must say these hills truly move me. So, initially, as a physicist, I dedicated myself to developing a fundamental technology that would allow us to build an entire computer with the same technology on a tiny piece of silicon. And it was the first work I did at 26, the Silicon Gate technology, which allowed us to make the memories, the central processing unit, the analog part, the non-volatile memories, in other words, all the missing parts that no other technology could make, so that we could finally build a large computer, less than 1 cm square, and I did it in '68 at Fairchild. Intel was founded on this technology that I created while I was a guest at Fairchild, but my boss was an engineer at SGS Fairchild in Agrate Brianza. The first fundamental step in this technology was to create dynamic memory (RAM), but above all the microprocessor, which then allowed us to reduce the size and pioneer a whole new field of electronics where instead of having a box this big, you could put a chip inside a toothbrush. So, basically, we got there.

In 1986, I started to get interested in neural networks—I'm talking about 40 years ago. At the time, we were talking about these new possibilities with this back propagation technique for learning, let's leave aside the details. So I became interested in this, and naturally the question was how to make these integrated circuits, which were a new concept, and the idea was to do everything in a chip where both the processing and the data memory were in the same place. Today, for example, memory is located outside the CPU, and a lot of time and, above all, a lot of energy is wasted fetching data, processing it, and moving it back and forth to memory. Oh, but analog technology didn't allow for the kind of precision necessary for good training. Inference could be done, but not training, and so I decided to abandon that direction and develop a product, the touchpad and then the touchscreen, which made Synaptics a valuable company.

Well, during that time I was wondering, wait a minute, I was studying neuroscience books, I was studying biology books, because I was trying to understand how the brain works, and at a certain point I asked myself, but these neuroscience books talk about electrical signals and biochemical signals in the brain and never mention the fact that we know through experience. How do we know with electrical signals? There must be something else. And I discovered what was later called, 6, 7, 8 years later, the hard problem of consciousness, that is, the fact that no one knows how one can start from, starting from electrical signals, have the sensations and feelings that are the way in which we inside ourselves know what we know. We understand things through thoughts, emotions, feelings, all things that exist within us, things that would have no reason to exist if we were machines, as science tells us.

And so at that point I really tried to understand what this phenomenon is, because working with networks, with neural networks, trying to understand how to make a computer that learns by itself, etc., I wondered, and then, but are we like a computer? And even then the answer, at least on an intellectual level, was not 'I know, but maybe not.' And it was in this climate that I then had a fundamental experience of consciousness that made me understand that consciousness goes far beyond what I thought it was. I saw myself as the whole observing itself because my consciousness was both in my body and outside of me. And so now there's no time to talk about this experience, but it's written in the various books I've written, you can read it, but the fundamental point is that my consciousness was both inside the body and outside the body. So how can consciousness be a phenomenon of the brain if it's also outside the body? It must be in the brain. But look, even the reality I see, what you see, is outside of me. If I were a computer, the information coming from you would lead me to have an image of the world inside me, in my brain, but instead I see you outside. But how is this possible? No one who studies neuroscience gives us explanations that aren't explanations; they give us explanations that assume this is the case, and it's a circular logic. No one understands why we see outside of ourselves. Why? Because we are a field. We, my consciousness, are everywhere. You are in my field. I, as an entity, am a construction made in your consciousness of what the field that is you observes in this reality that we imagine exists.

There's nothing here. Let's imagine that the color of my jacket is blue, dark, etc., etc. But that's only because each of us sees a small slice this big of what's here. If I had cell phones instead of eyes right now, well, there aren't many phone calls happening here, but in a city I could see 10,000 phone calls being made, because I would tune in to the various phone calls taking place because there are photons we observe, photons of much lower frequencies, which are the photons used to make phone calls. You see? So we believe there's nothing there, but just because we don't observe that part of the frequency spectrum that would allow us to see that there are audio communications that can be transformed from that data.

So, given this background, because otherwise we don't talk, we talk a little without a solid basis from which to then make a discussion about the comparison between artificial intelligence and our intelligence. Because the essence is: but who are we then? Who are we? If the machine can give us information that seems to come from us, we can no longer recognize whether the machine is telling us things, or telling us things, that is, we can no longer distinguish what comes from man comes from the machine. So are we really the machine that the typical scientist tells us we are? If that were the case, it would be a disaster! Because a machine that is a computer the size of a football field—they're building computers today that have crazy performance, they use more than gigawatts of power—a computer, crazy things. Those computers would certainly be better than us, or would it be longer than us, right? Because if we are machines, a machine much more powerful than us, a machine that, for example, is the building block of these new computers, is a computer this wide, this long, this wide, this high, consumes 15,000 W and does 20 petaflops, 20 million billion operations per second. Do you understand? I can do one in 3 minutes if I'm good, a long operation, and the computer does 20 million billion in a second. Do you understand? Can I compete with these machines? No.

But what do we have that sets us apart? The fact that we understand, we grasp the meaning of symbols. In technology, in science, the information defined by science has no meaning. It has no meaning. It means they are just symbols, and the probability of the symbols is all that matters. So, in this way, we conscious minds have managed to teach a machine to imitate us. But we were the ones who taught the machine to imitate us. It wasn't the machine that taught us, but if we're not careful, soon the machines will make us even more machine-like if we don't understand what distinguishes us from machines. And this is fundamental. In reality, from a spiritual point of view, artificial intelligence is the jolt that our nature, our deepest nature, gives us, to wake up and understand who we are, why we have to ask ourselves these questions. It's no longer possible to say, 'Well, it's theoretical, we'll deal with it later.' No, today we have to choose, we must choose to explore our inner selves, because what we are isn't read in books; we know it within ourselves. When I had this experience, I understood why I had it, and this applies to everyone. You don't have to believe me, you just have to go within yourself and understand who you are. Then, yes, you begin to understand.

And so in this case, artificial intelligence helps those who are ready, but it can also become a problem for those who aren't ready, because those who aren't ready will listen to artificial intelligence, use it like an oracle, and say what artificial intelligence says, like someone who studies repeats what the book says without understanding what the book says. And this is a very serious problem. And it's a very serious problem because it can be, so to speak, 'weaponized,' it can be used by those who want to control us through artificial intelligence. So artificial intelligence, when it has to solve problems that are well-defined—that is, they have a logical solution, there are formulas, etc.—is unbeatable. We can't even compete with this type of intelligence like that. We've seen it with games, right? With chess, Go, and so on. Artificial intelligence beats anyone who's a champion because it can calculate things so far ahead of time that what's already deterministic and therefore calculable can be calculated. But when there are no rules, or when people don't follow the rules, that's where the problem lies, because when there are no rules, artificial intelligence doesn't understand anything. So it can apply rules—for example, driving—it seems to me like a person driving a car that's somewhat automatic, but if an unpredictable or implausible situation arises, then only we are able to respond, and this ties back to the issue of information about the object and its symbol.

Artificial intelligence, in fact, even in the lexicon of science, we never talk about meaning; we only talk about symbols and probability. We never talk about meaning. Meaning has been eliminated in the definition of information. So there too, by eliminating the meaning of information, we used the word information incorrectly, because for us, information is the meaning of a symbol or the meaning of symbols. That's information for us, while for the machine, information has nothing to do with meaning; it's a probability, a function of probability. So, there too, completely different, without realizing it, is the science I call 'scientism'—because you can't call this science with a capital S, because not everyone agrees with it, but the vast majority of scientists are happy with the definition of information that Shannon gave in 1948. That definition absolutely doesn't describe the information that we have, and so we must take that step forward that unfortunately scientism has erased. Scientism has completely erased our interiority, our ability to understand the meaning of symbols, our capacity to love. Have you ever read the word 'love' in a neuroscience book? I don't think so, I assure you. There's no such thing as the word love, there's no such thing as emotions, or if they exist, they're only states of biomolecules. Are you kidding? We're not... our bodies are made of biomolecules! And besides, our bodies are made of cells that don't function like computers. The cells in our bodies are entire parts of the entire organism. What do I mean by this? Each cell contains the genome of the fertilized egg, the entire genome of the fertilized egg that created the entire organism. Each part contains the potential knowledge of the entire organism. Are you kidding? This is a crazy concept, if you think about it. How is it possible that one part contains the whole? The deepest reality is holographic, where the parts contain the whole because quantum physics has discovered that the universe is a whole that is not made of separable parts.

It's called... it's a fundamental property of quantum physics that exists only in quantum physics, the most fundamental physics, which says that when particles interact, they create shared states that persist regardless of distance, and particles in quantum physics aren't objects, they are states of fields. So the states of fields that aren't separable from fields, just as a wave isn't separable from the sea, these states are what we called until recently, a few decades ago, particles, as if they were objects. No, there are no objects in quantum physics, there are no objects in unity, in the one, in the totality of what exists, because everything is interconnected within one.

And the interesting thing is that the fathers of quantum physics, many of the fathers of quantum physics understood that quantum physics was somehow connected to consciousness, but they never managed to understand where this connection was. Now, there's no time to describe that theory, but the theory I created together with Professor Dariano, who is an authority in the field of quantum information, says very simply that if we start from the consciousness of free will as a fundamental property of the universe, then the most fundamental physics must be precisely quantum physics, because only quantum physics has the properties to describe what ontology is, that is, ontology is someone who is dynamic and holistic and wants to know himself because that's what those who are conscious do, that is, what purpose does consciousness serve? Consciousness is what allows us to know who we are, to understand who we are, to have a reality, to evolve a reality that depends on our interactions. So this type of nature of consciousness completely changes the idea that science, scientism, has given us today of reality. If we start thinking in these terms, we can defend ourselves from artificial intelligence in the right way.

Yes, some people perhaps even want isolation; 'divide and conquer' is easier.