Overview: Who This Is

Along one edge of the Intel 4004, the engineer who designed its circuits left two microscopic letters: F.F. Federico Faggin’s initials tell us something the rest of the chip cannot. We know who put them there and what he meant by them. The processor could follow an instruction, but nowhere in its design did it have to recognize its maker’s name.

Faggin, an Italian-born physicist and engineer, helped turn a Japanese calculator project into the first commercially available general-purpose microprocessor on a single chip. Years later, working on machines that could recognize patterns, he began asking a question he had once expected engineering to answer: could a machine actually experience what it recognizes? In his own example, a machine identifies a rose by its chemical signature. Has anything, anywhere inside the machine, smelled a rose?

His answer is that running a program is not enough. His further claim, developed with physicist Giacomo Mauro D’Ariano, is that experience belongs to the quantum foundations of nature. The distance between those two claims is where Faggin’s story gets interesting.

Faggin points to a detail of the Intel 4004 in 2011. Magnified to this scale, he can trace the layout with a finger. In the original, roughly 2,300 transistors fit on a single chip. Photograph: Intel Free Press, CC BY-SA 2.0

Origins and Background

Faggin grew up in Vicenza, Italy, earned a physics degree at the University of Padua in 1965 and moved to California in 1968 to work at Fairchild Semiconductor. There, the practical problem was how to fit more electronic switches onto a chip and get them to work reliably. A transistor has a gate that controls the flow of current. Faggin and colleague Tom Klein helped develop a manufacturing process that used silicon for that gate and used the gate itself to align adjoining parts of the switch. Faggin put the process to work in the Fairchild 3708, the first commercial integrated circuit to use it.

That background explains why Intel needed him. The Japanese computer museum’s account of the Busicom calculator records the customer’s problem: different calculator models called for different circuitry. What if the manufacturer could change a set of instructions instead of redesigning the hardware each time? By the time Faggin joined Intel in 1970, there was a plan for a programmable processor. There still had to be a way to build it.

What He Is Known For

A calculator that could have been something else

Intel’s Ted Hoff and Stanley Mazor worked out the broad processor design. Faggin led the work of turning it into silicon, with Masatoshi Shima of Busicom helping develop and check the detailed logic. The Computer History Museum’s account credits all four. The 4004 packed about 2,300 transistors into its central processing unit; companion chips supplied memory and other functions. What fit on one chip was the part that carried out instructions, not an entire working computer.

The distinction paid off. Busicom got its printing calculator. After the companies revised their agreement, Intel advertised the chip family in November 1971 for use beyond that original job. A designer could give the same processor different work by writing a different program. Faggin later cofounded Zilog with Ralph Ungermann and led development of the Z80, a processor that found its way into home computers. He had helped make a remarkable engineering principle ordinary: a machine’s material need not determine its task.

The question that followed him out of the laboratory

In 1986 Faggin cofounded Synaptics with Carver Mead. Its early researchers built circuits that could recognize patterns; the company put that expertise to work in laptop touchpads. A pad can locate a moving fingertip well enough to guide a cursor. The person using it also feels a surface under the finger. Only the first achievement appears in the device specification.

In an excerpt from his memoir Silicon, Faggin recalls initially thinking a conscious computer should be possible if the brain was a machine. During a holiday at Lake Tahoe in December 1990, he says, he woke in the night with an overwhelming feeling of love and of unity with the world around him. In a later interview, he describes the event as changing what he thought was fundamental. For Faggin it carried the force of discovery. For a reader it offers something different and valuable: his account of why he changed his mind.

In 2011 Federico and Elvia Faggin set up a foundation to support the study of consciousness. A 2015 gift matched by the University of California established a $1 million endowed chair at UC Santa Cruz in the physics of information. Its remit is wider than Faggin’s own theory. One tangible consequence of his change of direction was support for questions other researchers could pursue on their own terms.

The Core Idea or Signal

Return to the rose. An instrument could detect airborne molecules, compare the pattern with a library of samples and print the right word. You could check its accuracy bottle by bottle. What you cannot read from the printout is whether the successful identification smelled like anything to the device. Faggin calls attention to that difference between a symbol other people can read and a quality someone experiences. The technical term for such a felt quality is a quale; the plural is qualia.

D’Ariano, a physicist at the University of Pavia, and Faggin posted their proposal in December 2020 and published it as a book chapter in March 2022. They start with a strong assumption: the privacy of experience is a principle of nature. I can describe a scent to you, but I cannot hand you my moment of smelling it. Digital information behaves differently. A file can be copied, bit for bit, without anything missing from the duplicate.

Here quantum physics offers them a precise comparison. Its no-cloning theorem says there is no universal procedure that takes an arbitrary unknown quantum state and makes a perfect second copy while retaining the original. It does not mean that all information about a quantum system is inaccessible, or that two systems cannot be prepared alike when the state is known.

The authors propose that a certain kind of quantum state, described from outside by physics, is lived from within as an experience. They call their view quantum-information panpsychism: experience, on this account, is a basic feature of nature, already there before brains evolved. They suggest that an entangled state, in which the whole cannot be described as a collection of independent states for its parts, offers a way to think about the unity of experience. The scent, color and shape can be present to one person as a single rose.

This is why a better program, on Faggin’s account, would not by itself give the detector a sense of smell. It would improve what the machine does with the signal; his theory asks for a physical state with a point of view of its own. The copying theorem supplies a constraint on what an outside observer can do. It says nothing by itself about what, if anything, it feels like to be the system.

Two plots show an electronic sensor’s response over time and at increasing concentrations of natural rose oil.
In 2018, Minju Lee and colleagues measured an electronic sensor’s response to natural rose oil. As the concentration increased, so did the signal. The researchers could measure what the device detected. Faggin’s question is whether the scent was present to anything. Figure 5, Lee et al., Scientific Reports, CC BY 4.0.

Perspectives and Interpretations

The most revealing disagreement is with a philosopher Faggin cites. David Chalmers formulated the “hard problem” in 1995: explaining how a system discriminates, remembers and reports still leaves the question of why any of it feels like something. But in the same paper, Chalmers argues that if silicon components could replace neurons while preserving all their fine-grained causal relationships, experience should be preserved too. Imagine making the change one neuron at a time. At what replacement would the rose begin to lose its scent, if everything that allows the person to notice and compare the scent is still working? Faggin and Chalmers agree about what needs explaining. They disagree about what a faithful replacement would have to preserve.

Researchers Patrick Butlin, Robert Long and colleagues approach machines from another direction. Their work on indicators of AI consciousness asks what internal features would count as evidence under particular theories. One example is whether information can be made available across a system for memory, reasoning and action. This gives an investigator something more informative to inspect than a chatbot’s declaration that it feels. The authors treat the indicators as evidence of varying strength, not a certificate of sentience. Their approach and Faggin’s can agree that a fluent sentence settles little while disagreeing about what to examine next.

Strengths and Limitations

Faggin and D’Ariano do offer ways their proposal might meet an experiment. In section 7 of their paper, they consider drawings and stereograms that admit incompatible views. Look at a wire-frame Necker cube and the face that seemed nearest can suddenly seem farthest away. The lines have stayed put. What you see has changed. The authors ask whether such reversals could reveal something distinctively quantum, and acknowledge how difficult a decisive measurement would be.

The cube already comes with a useful complication. In a 2010 experiment with rotating wire-frame cubes, Sarah Harrison and Benjamin Backus trained observers with depth cues. Later, when the cues were removed, where the cube fell on the retina helped predict which way it appeared to rotate. That study did not test Faggin’s theory. It shows why seeing a reversal is insufficient: learned expectations can influence the result. A stronger test would specify in advance a pattern the quantum account predicts and a competing model of perception cannot reproduce. Even success there would establish a role in perception before it established that a quantum state is the feeling.

There is also a wrinkle in the copying argument. Suppose two physically separate machines did have identical experiences of a scent. Each experience could still be private: neither machine would thereby smell through the other one’s nose. Perfect copying and first-person privacy need not be opposites in the way Faggin’s argument requires. Free will poses a related difficulty. The authors interpret a particular quantum outcome as a conscious agent’s choice, but an event can be impossible to predict without having been chosen for a reason. The mathematics alone does not tell us whose reason made the difference.

Two offset squares form a transparent wire-frame cube; either square can appear to be in front.
Which face of the cube looks closer? Look a moment longer and the other face may come forward, though no line has moved. Giacomo Mauro D’Ariano and Faggin ask whether such reversals could help test their theory. The reversal itself does not establish a quantum cause. Diagram: BenFrantzDale, CC BY-SA 3.0;

Broader Implications

Faggin has taken the idea well beyond a technical readership. In the English-language book Irreducible (2024) and in interviews, he presents a world in which consciousness and agency are basic features of reality. He uses the word seity for an entity with an inner life, an identity and the capacity to choose. In a 2025 interview he makes a revealing concession: the mathematics he draws on can represent states, but in his view it cannot yet explain what those states mean to anyone. His Beyond the Invisible, published in English in October 2026, grew out of questions from his audience. That is one place his idea demonstrably traveled: from a technical collaboration into a public conversation about meaning, agency and spiritual experience.

There is a practical consequence even for someone who rejects his philosophy. A machine that correctly identifies the scent of a rose may be an excellent instrument. Whether it could suffer is a different sort of question, with consequences for how we ought to treat it. A performance test designed for the first purpose was never designed to settle the second. Faggin’s work asks anyone making that leap to say what further evidence warrants it.

The Reality Signal

What This Subject Represents

The same engineer who made it easier for machines to take on new tasks spent his later life asking what the ability to perform a task leaves out. He understood the power of describing a device by what it does; he also understood that a specification only answers the question written into it.

What Reality Frame It Challenges

A correct report is often good evidence that someone perceived something. Faggin challenges the assumption that the fit between report and experience must hold for any system producing the same words. His rose earns that caution. His quantum account proposes one way to explain the difference.

Why It Matters Now

A person and a language model can both produce the sentence “I’m in pain.” In the human case we have more to go on than the sentence: a body, a history, and other people whose pain we recognize in light of our own. As machines acquire the language of inner life, the grounds for extending that recognition deserve to be made explicit.

What Remains Unresolved

No result discussed here identifies a particular quantum state as the experience of a particular scent. Nor do we know whether preserving everything a nervous system does would preserve everything its owner feels. Those are distinct open questions; settling one would change how we investigate the other.

Pierre Joseph Redouté’s botanical engraving of a pink Frankfort rose with buds and green leaves.
Pierre Joseph Redouté’s Frankfort rose, from Les Roses (1817–24). The engraving gives us enough detail to recognize the flower. What it cannot give us is the scent itself. Image: The Metropolitan Museum of Art, public domain.

The Galactic Mind Perspective

The part of Faggin’s argument we would keep even if his theory were abandoned is his habit of asking exactly what a test has shown. Give his imagined detector a hundred unlabeled samples and it might earn our confidence a hundred times over. Its success would be real. We could send it into a greenhouse to identify varieties we ourselves cannot tell apart, and still have learned nothing decisive about whether it enjoys the visit.

We give less weight to the proposed identity between experience and a quantum state. The obstacle is specific: a proof that an unknown state cannot be copied still leaves unexplained why one state has the character of a fragrance and another the sting of pain. Faggin’s strongest contribution here is the demand for a theory that can tell those experiences apart and explain why the difference matters to the one having them.

Open Thread

What would you need to learn about the rose detector, beyond every answer it gives, to have a good reason to believe it smells the flower?

More in Dossier

Sources / Receipts