NO, Artificial Intelligence is not what you are told
Jacques Mojsilovic
Chief Communication, Marketing and Partnerships Officer at Numalis - The No Guess Company - Trustworthy and Responsible AI advocate
OpenAI's latest AI model, the "Strawberry project," is generating buzz with claims of "human-like" reasoning capabilities. We’ve seen countless AI announcements like this lately, and frankly, I’m tired of the abusive and misleading and dangerous language.
#Mythbusting: NO, AI cannot reason, here's why.
Reasoning is the cognitive process of thinking about something in a logical way in order to form a conclusion or judgment. It involves the use of evidence, facts, and principles to draw inferences or arrive at decisions. It can be deductive (deriving specific conclusions from general principles) or inductive (inferring general conclusions from specific instances).
Etymology of Reasoning:
Latin Origin (circa 13th century): The word "reasoning" stems from the Latin root ratio, meaning “reckoning, understanding, reason.” It relates to reri, which means “to think, consider, or reckon.”
Middle English (14th century): The term reson in Middle English was borrowed from Old French raison, which had a similar meaning of “intellectual faculty, reason, judgment, or justification.” At this time, it was closely tied to the notion of logical argumentation and thought.
Enlightenment Period (17th-18th centuries): During the Enlightenment, reasoning became a central theme in philosophy, as thinkers like Descartes, Locke, and Kant emphasized reason as the primary source of knowledge. This period also saw a shift towards systematic scientific reasoning, differentiating between deductive and inductive methods.
Modern Usage (19th century-present): The term "reasoning" evolved to include more psychological and scientific connotations, reflecting advances in logic, cognitive psychology, and the philosophy of science. Today, reasoning encompasses a broader scope, including problem-solving, decision-making, and artificial intelligence (AI) applications.
Reasoning, a fundamental cognitive process rooted in our intellectual history, has evolved to encompass a wide range of applications, from philosophical inquiry to artificial intelligence. While the term "reasoning" often implies human-like capabilities, it is essential to distinguish between the algorithmic processes employed by AI and the complex cognitive functions of the human mind. Anthropomorphizing AI not only leads to misconceptions about its capabilities and limitations but also risks distorting societal understanding and ethical considerations.
Psychologically, attributing human-like reasoning to AI can foster over-reliance on technology, undermining critical thinking and personal responsibility. Scientifically, such confusion blurs the line between genuine cognitive processes and automated data processing, where AI excels at pattern recognition but lacks consciousness, intention, and self-awareness. Societally, these misunderstandings could influence policy and public perception in ways that fail to account for AI’s actual limitations and strengths.
领英推荐
To draw an analogy: before the digital age, we had physical mail, which evolved into e-mail, a faster and more efficient but fundamentally different mode of communication. Similarly, we might think of AI as performing e-reasoning, a process that mimics some aspects of human reasoning but operates on entirely different principles. Just as we adapted our language to distinguish between mail and e-mail, we must develop nuanced terminology to describe AI processes accurately, ensuring clarity as technology continues to advance.
Education is paramount and it starts at school, not on social media, neither on Linkedin, not even with so called "AI experts" influencers not able to use proper terminology.