Can a computer really taste chicken?

In this blog post, starting with a scene from ‘The Matrix’, I’ll examine the process by which humans perceive taste, the development of artificial taste, and John Searle’s “Chinese Room” thought experiment to consider whether a computer can go beyond simply recognizing taste to actually tasting and understanding the flavor of chicken.

 

‘The Matrix’ is a film that depicts humans forced to live in a virtual reality called “the Matrix” by machines with artificial intelligence, as well as their resistance. While the virtual reality of the Matrix mirrors the everyday life of an ordinary person in 1999, the reality outside the Matrix is bleak beyond compare. The members of the Resistance all have haggard faces, are clad in ragged clothes, and subsist on a pale, insipid porridge whose flavor they can’t even identify. As the film unfolds, there’s a scene where a member of the Resistance is eating on a spaceship; one of them complains about their tasteless food while extolling how much better the chicken they tasted inside the Matrix was compared to that porridge. While this scene foreshadows the upcoming appearance of a traitor who wishes to return to the Matrix, it also raises an intriguing question. In a previous scene, when one person voiced dissatisfaction with their meal, another questioned whether the Matrix’s artificial intelligence could actually understand taste, suggesting that the taste of the chicken they had eaten and experienced within the Matrix might not have been the true taste of real chicken. Can the Matrix—or, in other words, a computer—truly perceive the taste that humans experience?
To answer this question, we must first understand how humans perceive taste. Humans perceive taste through the taste receptor system in the mouth and throat, including the tongue, which is the primary organ of taste. Taste is formed when chemicals in food stimulate taste cells within the taste buds. Currently, there are five generally recognized basic tastes: sweet, bitter, sour, salty, and umami. Taste cells react to chemicals associated with each taste, causing changes in ions or signaling molecules, which in turn generate electrical signals. These signals are transmitted to the brain via the facial nerve, the glossopharyngeal nerve, and the vagus nerve, which are distributed throughout the tongue and pharynx.
For a computer to perceive taste, this process must be implemented mechanically; in particular, to correlate the results with human taste experiences, it is necessary to appropriately generate and process signals corresponding to sensory stimuli. Currently, in the fields of artificial vision and hearing, devices that detect external stimuli have advanced significantly, and this level of advancement can be easily observed through high-performance cameras and microphones. In contrast, due to its inherent complexity and precision, artificial taste still faces challenges in faithfully reproducing human taste. However, this does not mean that research on artificial taste remains in its early stages. Artificial taste sensors, known as “e-tongues,” have already been developed and are being used to analyze characteristics related to sweet, sour, salty, bitter, and umami tastes using various chemical sensors, as well as to evaluate the taste of foods and pharmaceuticals. As of 2026, research is also underway on multisensory e-tongues designed to detect not only chemical properties but also physical characteristics such as temperature and texture. Therefore, we can now definitively answer “yes” to the question, “Does a device exist that allows a computer to perceive taste?” However, whether such a device can produce a subjective taste experience similar to that of a human is an entirely different matter.
Yet a more fundamental question remains: Just because a computer can receive certain stimuli and appears to respond correctly to them, does that mean it actually understands them? Suppose a computer eats chicken and an artificial taste device generates the appropriate electrical signals to recognize it. Can the computer actually conjure up the images that come to mind when we eat chicken—not only the chicken itself but also memories associated with it and various images evoked by its taste—and “understand” them? Can a computer, based on the electrical signals received through its artificial taste senses, imagine anything beyond the chicken it has eaten?
Let’s return to the movie for a moment. In the latter part of the film, the protagonist storms into the building occupied by their enemy computer programs to rescue a colleague who has been captured by them. There is a scene where one of the enemies is interrogating the protagonist’s colleague, and a memorable line is delivered. “I hate this place. This zoo, this prison, this reality—whatever you want to call it—I can’t stand it anymore. It’s because of the smell—if such a thing even exists. I feel like that smell is seeping into me too. I can smell your foul stench. And every time that happens, I get scared that I’ve been infected too.”
The program that featured this dialogue expresses revulsion toward the smell of humans, yet interestingly, it adds the qualifier “if such a thing exists.” Based on this, it seems the program is aware of the concept of “smell” but does not fully grasp the abstract concepts it refers to. Some argue that the fact the program describes “tasting” a foul stink indicates it cannot even distinguish between images derived from the sense of smell and those from the sense of taste. Setting aside the fact that the expression “tasting a smell” is sufficiently ambiguous, it remains questionable whether the program in the film understands the sensory concept of smell in the same way humans do. Although he acts just like a real person, he appears to be merely following a pre-programmed sequence and is incapable of anything beyond that. Even if he provides appropriate responses to metaphysical questions, it seems difficult to say that he truly understands them, as he is simply following a predetermined sequence.
There is a famous thought experiment in the field of artificial intelligence related to this: John Searle’s “Chinese Room.” Searle posited a room containing a person who can speak only English. This person is given a writing tool, a pre-prepared list of questions in Chinese, and a complete list of corresponding answers. Now, if a Chinese examiner writes a question in Chinese and slips it into the room, the person inside will write the appropriate answer based on the prepared list of responses and hand it to the examiner outside. From the Chinese examiner’s perspective, since the person inside the room provides the correct answers to their questions, it would be reasonable to assume that he or she is fluent in Chinese. However, the person inside the room has no idea what is being said to them or what they are writing in response. Using the Chinese Room, Selle argued that the mere fact that a system produces the appropriate output for a given input does not establish that the artificial intelligence possesses understanding. The Chinese Room thought experiment was first proposed in Selle’s 1980 paper “Minds, Brains, and Programs” and has since long been a subject of debate in the fields of artificial intelligence and the philosophy of mind.
Applying this to the computer with artificial taste mentioned above, we can conclude that the computer merely knows what it has eaten or which chemical signals correspond to the characteristics of a particular food; it cannot be said to actually “taste” it. In other words, while the computer can correctly answer a few predetermined questions, that is the extent of its ability—and even then, it is not answering based on understanding but merely parroting predetermined responses. However, this raises several questions. First, does the fact that the person in the room does not understand Chinese constitute a valid argument against the existence of intelligence? Consider the human brain as a counterexample. The human brain is composed of countless neurons. Each of these cells performs only the basic function of storing data or transmitting information. To use a computer analogy, it would be comparable to a small computer with only simple functions. However, if we follow Seol’s argument, since each of these cells merely receives specific electrical signals as predetermined and outputs appropriate electrical signals in response, brain cells are incapable of the act of understanding, and consequently, the human brain—as the aggregate of these cells—is also incapable of it. Of course, this is a conclusion that is difficult to accept. In response, Seol hypothesized a scenario involving multiple water pipes corresponding to neurons and valves that regulate them, with a person who adjusts these pipes appropriately based on given questions and answers to produce results. He then countered by asking where understanding could possibly arise within this system, given that the person adjusting the pipes still does not understand Chinese.
However, this system is merely a modified version of the Chinese Room argument presented earlier. Whether it is the Chinese Room or the water pipe building, there is no difference between them as systems with states and inputs and outputs—in other words, the former consists of a pen, paper, and the characters written on it, while the latter consists of valves, water pipes, water, and the person operating them. The claim that neither the person nor the water pipes in the water pipe building understand Chinese is the same as saying that neither the person in the Chinese Room nor the pen or paper they use understands Chinese. In other words, just as the Chinese Room does not serve as a valid refutation of the human brain, neither does the water-pipe building. This applies equally to other refutations of the theory.
There is one more question to consider. Suppose technology advances to the point where an entire human consciousness can be transferred to a computer. Naturally, the computer storing human consciousness would not operate according to biochemical principles like the human brain, but rather according to electrical principles. This human consciousness would produce an output specified in the consciousness data in response to any input received by the computer; therefore, it could also be viewed as a type of program. However, according to Searle’s argument, this too cannot be considered a system capable of human understanding—simply because the storage medium has changed while the brain’s entire input-output structure remains unchanged. Thus, the mere existence of predetermined input-output processes is insufficient to deny the existence of understanding, let alone intelligence. While Searle’s “Chinese Room” experiment may serve as a rebuttal to the claim—as Searle originally intended—that passing the “Turing test” is equivalent to the subject possessing intelligence, it does not prove anything beyond that. The Turing Test, too, is widely discussed today as a historical concept for evaluating the capabilities of artificial intelligence and human-like conversational behavior; merely passing the test does not prove the existence of consciousness or subjective experience.
Just as with the Chinese Room experiment mentioned earlier, while some who pass the Turing Test may possess intelligence, it cannot be said that all who pass do; similarly, just because some who pass lack intelligence does not mean that all of them do. There is considerable room for debate regarding the claim that all forms of artificial intelligence or intelligence can be dismissed based solely on this argument. While the Chinese Room argument raises a powerful question about whether human-like understanding or consciousness can be established merely through the execution of a computer program, it has not put an end to all discussions regarding the nature of intelligence and understanding. In fact, since its introduction, the Chinese Room argument has sparked a long-standing debate among various philosophers and cognitive scientists, giving rise to counterarguments and rebuttals.
Starting with a scene from ‘The Matrix’, we have explored the principles of human taste perception, the development of artificial taste, and the arguments surrounding the Chinese Room experiment to examine whether computers can not only perceive taste but also truly experience and understand it. In the past, devices designed to simulate artificial taste were very limited, but today, artificial taste sensors—such as electronic tongues—actually exist and are used to analyze the ingredients and flavors of foods and pharmaceuticals. However, these devices have not yet reached the stage where we can say they perceive the subjective taste experienced by humans exactly as it is. Therefore, it remains difficult to answer the question, “Can a computer taste chicken?” with a simple “yes” or “no.”
Nevertheless, this does not necessarily negate the possibility of a computer that can truly “taste chicken.” Just as with human neural networks, while individual components may be nothing more than simple input-output devices, if they form a sufficiently complex structure—even if not strictly in the form of a neural network—and possess sufficient learning capabilities, we cannot completely rule out the possibility that they might exhibit intelligence similar to what humans understand. Of course, there is no evidence that current artificial intelligence or electronic tongues possess the same subjective taste experiences as humans. However, at some point in the future—when technology has advanced enough to successfully implement artificial neural networks, sensory sensors, and learning systems, and our understanding of human consciousness and experience has deepened—we may find ourselves rewatching ‘The Matrix’, which will by then have become a classic masterpiece, and answering the question posed in the introduction with, “It wasn’t true before, but it is now!”

 

About the author

Cam Tien

I love things that are gentle and cute. I love dogs, cats, and flowers because they make me happy. I also enjoy eating and traveling to discover new things. Besides that, I like to lie back, take in the scenery, and relax to enjoy life.