ARTICLE UPDATED MAY 2026
The border between humans and machines: from fiction to the laboratory
Defining the boundary between robot and human has been one of the major themes in literature and film. Asimov and his “Three Laws of Robotics,” the replicants from Blade Runner, or Mary Shelley's “Frankenstein” explore, through fiction, what happens when a technological creation becomes too much like us.
Today, that debate is no longer just narrative: advances in generative artificial intelligence and in agents capable of making decisions reopen questions about consciousness, responsibility, and control. Long before all this existed, Alan Turing proposed a simple experiment to reframe the question “Can machines think?” in observable terms.
What is the Turing test?
The Turing test is an experiment that attempts to measure the extent to which a machine's behaviour can be indistinguishable from that of a human. In its classic formulation, an evaluator holds a written conversation in natural language with two hidden interlocutors: one is a person and the other is a machine.
If, after a reasonable length of conversation, the evaluator cannot reliably distinguish which is the machine, the machine is said to have “passed” the Turing test. Turing called this experiment the “imitation game” and proposed it as a practical way of avoiding abstract debates about definitions of “thought” or “intelligence”.
Over the decades, competitions and variants of the test have been organised, in which different programmes attempt to convince human judges that they are real people. One of the best-known cases was that of the chatbot “Eugene Goostman”, which in 2014 managed to persuade approximately one third of the judges that it was speaking to a human teenager, although the conditions of the test and its interpretation generated debate.
Nowadays, with the arrival of advanced language models, new experiments have emerged claiming that certain AIs have passed more demanding versions of the test, which has reignited the discussion about their real utility as a measure of intelligence.
Advantages of the Turing Test in artificial intelligence
Although it is now regarded as limited, the Turing test has several merits that explain why it remains so influential.
- Practical simplicity
Instead of trying to define complicated concepts like “thinking” or “understanding,” the test focuses on something very concrete: whether a machine can hold a conversation that a human would mistake for another person’s. This simplicity makes it easy to explain and implement, even outside of academic settings. - Thematic diversity
The test does not restrict the topics of conversation, so the system must be able to respond to a wide range of issues, from trivial questions to more abstract topics. This requires systems to handle a broad range of knowledge and types of dialogue, something that is particularly relevant in the era of generalist language models. - Social and empathetic dimension
Beyond the data, the test assesses the machine's ability to manage human nuances: ambiguities, humour, courtesy, or the minor inconsistencies typical of a real conversation. In a way, it also measures the AI's skill in adapting to the interlocutor and generating a sense of closeness, something very present in current conversational assistants.
Weaknesses and criticisms of the Turing Test
Over time, the Turing test has come in for a great deal of criticism and is now regarded, above all, as a historical milestone rather than a technical standard for evaluating AI systems.
One criticism is that the test does not measure deep understanding or genuine reasoning, but rather the ability to imitate. A machine can be optimised to appear human, including typographical errors or evasive responses, without this implying that it “understands” what it is saying. In fact, some programmes have used strategies such as pretending to be a child or someone with limited command of the language to justify inconsistent responses.
Another limitation is that the test does not detect behaviour that is far superior to that of humans. If a machine instantly solves extremely complex mathematical problems, the evaluator may suspect that it is a computer, even though such performance is, in itself, evidence of exceptional abilities. Paradoxically, a system that is “too intelligent” might fail the test if it does not adjust its behaviour to appear more human.
For these reasons, many researchers propose supplementing or replacing the Turing test with a series of tests focused on specific tasks: logical reasoning, reading comprehension, planning, multimodal interaction and ethical assessment, amongst others. The idea is to move away from a single, general experiment towards a set of metrics that better reflect the various dimensions of artificial intelligence.
Artificial intelligence as a career path in computer engineering
Beyond the discussion of the Turing test, artificial intelligence has become one of the most relevant career pathways within computer engineering. From language models capable of holding complex conversations to computer vision, recommendation, or planning systems, AI is applied today in sectors as diverse as health, finance, industry, marketing, or public administration.
Working in AI involves much more than trying to “pass a test”: it requires designing algorithms, managing data, evaluating models, considering ethical and safety aspects, and being able to take solutions from the lab to real-world products. Therefore, a solid foundation in programming, data structures, mathematics, and systems architecture remains fundamental for any professional wishing to specialise in this field.
If you’re interested in these topics and would like to get involved in the design, development and evaluation of artificial intelligence systems, the first step is a solid background in software engineering and computer science. At IMMUNE, you can explore our technology-related academic programmes and find the programme that best suits the area you wish to specialise in, whether that’s AI, software development, data or cybersecurity. The aim is to help you turn your interest in technology into a solid career, equipping you for the challenges and opportunities that the next generation of intelligent systems will bring.
