Artificial Intelligence is any device that perceives its environment and takes actions that maximize its chance of success at some goal.

I got this definition from Wikipedia that cited "Russell and Norvig (2003), Artificial Intelligence: A Modern Approach".

A transistor is a device that amplifies or switches electronic signals when it received an input signal.

Could one say the transistor is the AI?

It is certainly a basic building block of every AI out there, but is it an AI itself, albeit the most basic one?

I'm looking at it from a technological and economic point of view, leaving philosophy aside. From an economic perspective it seems to be an AI because transistor does useful work that it took an intelligent human to perform less than a century ago. And it does it completely on its own.

  • $\begingroup$ Welcome to our small brilliant community in small wires of the internet.when we say that a being or a program is self aware or has got self awareness,self reasoning/thinking that's inline and expected to beyond humanity.That's what artificial intelligence is all about.if you natural intelligence then artificial intelligence should be easy to define effectively.So do your transistors have self awareness or reason before performing a given task.or just automation powered. $\endgroup$ – quintumnia Aug 21 '17 at 10:41
  • 3
    $\begingroup$ And for your information,artificial intelligence is not a device..I disagree with that statement or definition,scientifically.That's a definition of a robot.not AI $\endgroup$ – quintumnia Aug 21 '17 at 10:53
  • $\begingroup$ defining what is AI on a case by case basis doesn't seem to be a very productive endeavor though. why care? $\endgroup$ – k.c. sayz 'k.c sayz' Aug 21 '17 at 15:18
  • 1
    $\begingroup$ I quite like this question! Welcome to AI. :) $\endgroup$ – DukeZhou Aug 21 '17 at 19:17
  • $\begingroup$ @quintumnia That's why I favor more precise terminology: "Agents" for AI in an economic (optimization) context and "Automata" for AI in a procedural context. There's no reason a mechanical AI could not be produced, and as nanotech advances, we might even see quite sophisticated mechanical AI. (There were certainly fairly sophisticated mechanical calculators.) A robot is just an automata in a physical, as opposed to a virtual, context. $\endgroup$ – DukeZhou Aug 21 '17 at 19:35

I think it might come down to whether the transistor is making a decision. If the transistor is being used as a switch, that would seem to qualify as a decision, even though it's an extremely rudimentary decision.

Intelligence, in reference to Artificial (or Algorithmic) Intelligence, is not restricted to high intelligence. A brute force Tic-Tac-Toe AI has extremely low, narrow intelligence but still constitute AI. An automatic switch, which makes the most simple decision possible, a binary choice, would seem to be the most basic form of intelligence.

Norvig's definition seems rooted in game theory, which is important in terms of utility of intelligence. But in a condition of intractability one is only assuming one's decision is more optimal than other choices. Outcomes can be evaluated, and a determination made as to whether the algorithm was "smart" or "dumb", but these terms refer to relative positions on a spectrum.

It's worth noting this fundamental definition of AI opens up a can of worms in that pre-transistor, automated mechanical switches such as the Strowger switch would also probably qualify.

And automata do not have to be electrical. The History of Automation wiki suggests the first feedback control system was for a water clock invented by Ctesibius. This device dates to the 3rd Century BCE, and water clocks were said to be the most accurate time pieces until Huygens.

This type of intelligence I tend to think of as autonomic, in the sense of involuntary, and distinct from higher functions, which are more commonly associated with "intelligence".

Note: The characteristics of an autonomic system in a computing context are quite interesting and include self-optimization, self-learning, and self-awareness.

ADDENDUM: After much thought, in reference to @JT's answer, I can't see how decisions can be separated from goals—there needs to be an intent, or the decision is merely random. This might prompt the question "can a simple switch be said to have a goal?"

  • $\begingroup$ ,so by basing on your ideology,transistors were the first to exhibit intelligence.All the answers here are based on an assumption. $\endgroup$ – quintumnia Aug 21 '17 at 20:57
  • $\begingroup$ Am going to do my research on this,after proving it right.then upvotes will rain on answers here. $\endgroup$ – quintumnia Aug 21 '17 at 21:05
  • $\begingroup$ @quintumnia More of a disclaimer. (There are so many areas of computing, and as a tech product manager and I try to have a good, basic knowledge of all of them. But on the flip side, being a "generalist" also means I'm not a specialist in any given area.) But, under this definition of AI, mechanical devices like the Strowger switch would probably also qualify. Answer updated. $\endgroup$ – DukeZhou Sep 11 '17 at 15:55
  • $\begingroup$ hope sometimes,we need to disapprove some scientific theories. $\endgroup$ – quintumnia Sep 11 '17 at 16:25
  • $\begingroup$ @quintumnia It's more about the definition of "Artificial Intelligence" which is a little fuzzy. Jayden Travnik's answer sheds light on the issue, and provides excellent context. $\endgroup$ – DukeZhou Sep 11 '17 at 16:29

Replacing my previous ill conceived answer with this definition of Intelligence from Richard Sutton (a founder and leader Reinforcement Learning) should answer your question.

John McCarthy long ago gave one of the best definitions: "Intelligence is the computational part of the ability to achieve goals in the world”. That is pretty straightforward and does not require a lot of explanation. It also allows for intelligence to be a matter of degree, and for intelligence to be of several varieties, which is as it should be. Thus a person, a thermostat, a chess-playing program, and a corporation all achieve goals to various degrees and in various senses. For those looking for some ultimate ‘true intelligence’, the lack of an absolute, binary definition is disappointing, but that is also as it should be.

The part that might benefit from explanation is what it means to achieve goals. What does it mean to have a goal? How can I tell if a system really has a goal rather than seems to? These questions seem deep and confusing until you realize that a system having a goal or not, despite the language, is not really a property of the system itself. It is in the relationship between the system and an observer. (In Dennett's words, it is a ‘stance’ that the observer take with respect to the system.)

What is it in the relationship between the system and the observer that makes it a goal-seeking system? It is that the system is most usefully understood (predicted, controlled) in terms of its outcomes rather than its mechanisms. Thus, for a home-owner a thermostat is most usefully understood in terms of its keeping the temperature constant, as achieving that outcome, as having that goal. But if i am an engineer designing a thermostat, or a repairman fixing one, then i need to understand it at a mechanistic level—and thus it does not have a goal. The thermostat does or does not have a goal depending of the observer. Another example is the person playing the chess computer. If I am a naive person, and a weaker player, I can best understand the computer as having the goal of beating me, of checkmating my king. But if I wrote the chess program (and it does not look very deep) I have a mechanistic way of understanding it that may be more useful for predicting and controlling it (and beating it).

Putting these two together, we can define intelligence concisely (though without much hope of being genuinely understood without further explanation): Intelligence is the computational part of the ability to achieve goals. A goal achieving system is one that is more usefully understood in terms of outcomes than in terms of mechanisms

  • $\begingroup$ So my understanding is that it needs to change the way it processes input by itself to be considered AI? $\endgroup$ – Arthur Tarasov Aug 21 '17 at 10:39
  • $\begingroup$ Not necessarily, rather it needs to have a goal and take actions towards that goal. In this way, a thermostat is very primitive AI as it has a goal of holding constant temperature. A transistor does not have any goal. $\endgroup$ – Jaden Travnik Aug 21 '17 at 10:44
  • $\begingroup$ @JadenTravnik ,Can you give us a hint on what is intelligence inline with your explanation.Tell the OP the truth instead of bringing here science fiction.The definition of AI is what is bringing messes here. $\endgroup$ – quintumnia Aug 21 '17 at 11:04
  • $\begingroup$ @JadenTravnik Isn't it similar to what a transistor does? Thermostat in a fridge turns on the cooling system when it receives the signal from thermometer that the temperature rose above threshold. A transistor can have a goal to turn something on when it receives the signal. $\endgroup$ – Arthur Tarasov Aug 21 '17 at 11:17
  • 2
    $\begingroup$ @ArthurTarasov That is a good point, I updated my answer with Rich Sutton's definition to elaborate and correct my first answer. $\endgroup$ – Jaden Travnik Aug 21 '17 at 13:13

Transistor is very similar in its function to single neuron and because of that one transistor could be considered to be a very tiny neural network - from this perspective it could be considered to be a form of artificial intelligence on their own. But transistor is not the first building block of AI, the first building blocks are the smallest particles that are possible in phisics - those are the building blocks of any machine. Also the definition is somehow hard to accept, as there is no reason why thermostat can be considered AI on their own and transistor not. Since thermostat needs an environment in which it works in a way that fulfills the definition of AI, so the transistor could be put in such environment - there's no difference. We consider that something is AI when it maximizes chnace of some goal, but who is defining this goal? We do. So if you say that the goal of some AI system is to have the output current equal to input current multiplied by factor of x, then you can say that bipolar junction transistor is the AI that does exactly that. Of course you can build a processor and memory from such transistors, then write software for eg. face recognition for a computer builded from them. It is all about connecting algorithms together to perform more sophisticated functions. The first "logical gates" (and builing blocks of AI) are the smallest particles possible, the smallest particles that can interact with each other to produce some output available to be used in next interaction - the whole universe works bacause of that - the matter just follows the algorithms - laws of phisics. On the side note: The only thing that for me is not algorithmical is consciousness, I think it is impossible to be created by just algorithms working together. Algorithms doesn't have any feelings and any number of them will not have feelings either - this will only process information without any consciousness inside. Intelligence is computational, consiciousness somehow reaches beyond phisical/algorithmical world - it is worth consideration - I personally think that we are like players in a game and there is other dimension/world from which our consciousness comes from, good question is: what are the rules of this game?


Could one say the transistor is the AI?

Correct. The single-electron transistor is a cellular automaton. Quantum dots are used to form a neuromorphic circuit, which is sometimes called a “nv-neuron”. Analog CMOS Implementation of a CNN-Based Locomotion Controller With Floating-Gate Devices It is the fundamental building block of any AI, can have feelings and a consciousness.


A transistor is definitely not the first artificial intelligence, for the following reasons.

  • Transistors are a device to vary the resistance through two P-N semiconductor junctions via an applied voltage at the inner layer, called the base, relative to one of the outer layers, called the emitter. Prior to transistors, vacuum tube triodes were a device to vary the conductance between a plate anode and a cathode. They both provide signal amplification and can be used to form logical and, or, nand, and nor gates and thus can be used to create flip-flop circuits and other computing devices. Tubes have a shorter life, greater size, and greater power consumption, so they have gone out of vogue for most circuit designs, but they were first.
  • The transistors used in typical VLSI circuitry deliberately do not remember data. They are stateless and therefore cannot learn or exercise the result of learning other than in the context of the circuit design.
  • Transistors, by themselves, cannot perform a calculation.
  • Transistors cannot make a decision based on the comparison of two options using some criteria for selection.

Transistors can, however, be used in larger designs to simulate or replicate forms of human behavior considered to be features of intelligence.

The definition, "Artificial Intelligence is any device that perceives its environment and takes actions that maximize its chance of success at some goal," quoted from Wikipedia's citing of Artificial Intelligence: A Modern Approach, Russell and Norvig, 2003, is inadequate for these reasons.

  • Perception is left undefined and is horribly ambiguous. Is it simply data input or comprehension of what is happening objectively in the environment? These are distant extremes.
  • Not all demonstrations of intelligence are in relation to success. Some would wisely point out that intelligence often transcends society's success measurements. A chess player that concedes long before onlookers see any sign of an impending loss, is an example. There may be a large number of brilliant conclusions that no one will ever see or hear because the thinker died poor and her or his intellectual output are buried in a garbage dump.
  • Goals and discoveries frequently arise from intelligent thoughts that are not maximized. Most of the roots of science were invented before the concept of maximization.
  • Thought experiments involve no external actions other than to communicate their result so others can benefit from an already completed intelligence process. (Ernst Mach, Einstein, Newton, Socrates, Gauss, Turing, Kurt Gödel, Lavoisier, Kekulé)

There is almost nothing that is not misleading in the Russell and Norvig definition, from either a technological or economic standpoint.

The statement, "From an economic perspective [the transistor] seems to be an AI because [a] transistor does useful work that it took an intelligent human to perform less than a century ago, and it does it completely on its own," doesn't stand up to scrutiny either.

  • Transistors don't do work on their own. They require input, at least one other component for biasing, sometimes another transistor and sometimes a resistor, and a load, which may or may not perform useful work.
  • Which is more intelligent, a switch, a person throwing the switch, the installer of the switch, the person that engineered the switch, or the person who devised a way to connect millions of electronic switches that could be programmed to simulate a multi-layer perceptron? The switch, whether a transistor or a vacuum tube or manual switch, is the probably the least intelligent of these.

Even if we compare a transistor to a biological neuron, they are miles apart in terms of informational complexity. Humans are still a long way from fully understanding how neuroplasticity, brain chemistry, and neural circuit topology relates to the features of adaptation, language, comprehension, cognition, and rational thought.

  • $\begingroup$ A simple memristor-network (which is a stateful device) can be realized with a Coherer invented 100 years ago. Secondly a “neuron transistor” behaves like a biological neuron. This is called a neuromorphic architecture. $\endgroup$ – Manuel Rodriguez Nov 9 '18 at 19:22
  • $\begingroup$ @ManuelRodriguez, memristors are not closely related to neuromorphic architectures. The choice of semiconductor technology and what can be morphed are independent aspects of circuit design. $\endgroup$ – han_nah_han_ Nov 9 '18 at 23:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.