A Whisper of Caution to the Ears of Intelligence

by: Mr. Antonio M. Cailao, FICD

ICD Fellow

Institute of Corporate Directors

Foreword:

The goal of Artificial Intelligence (AI) is to create systems and algorithms that can perform tasks that typically require human intelligence. This article entitled "A Whisper of Caution to the ears of intelligence” highlights the limitations of Artificial Intelligence. Such limitations are the lack of common sense, interpretability, and lack of many cognitive abilities.

---

One and half months ago, I wrote a piece vigorously urging a local bank to adopt an AI strategy and plan for BoC. Since then, the mania on the wonders of AI continues to escalate and sweep the world off its feet. Undeniably, I, too, am a fan but this AI fandom has its extreme members, unconditionally declaring AI as the “be all and end all”. But despite some signs of faltering, bizarre incidents created by generative AI — to name a few: autonomous vehicle accidents, unnecessary mastectomies for breast cancer patients, six fingers in one hand (!)— it does not begin to capture the potential gravity of what havoc AI can spawn. While to go beyond is as flawed as to fall short, amidst the cacophony of AI uproar, it now makes me want to retreat a bit and don my “moderation” cap. Hence, I wrote a restraining “pull back” piece which I want to share with you, a short essay entitled, “A whisper of caution to the ears of Intelligence”, a sobering cautionary tale of sorts highlighting the covert limitations of Artificial Intelligence. A whisper of caution to the ears of Intelligence… Artificial Intelligence uses billions of data and examples to train a software model, the structure of which is loosely based on the neural architecture of the brain. The resulting systems can perform myriad applications —such as recognizing images or speech, processing large language models, creating smart content, automating countless tasks and more besides—far more reliably than expert technicians formulate highly hand-crafted models… but they are not “intelligent” nor sentient in the way that most people understand the term, i.e., the capacity to experience sensations, feelings, and consciousness. They are powerful pattern-recognition tools learned during training but lack many cognitive abilities that even a child’s biological brain takes for granted in aspects of simple human cognition that involves self- awareness, subjective experiences, and Heartfelt emotions. AI can atypically struggle when asked to provide reasons—routinely generalizing from the rules it has earlier discovered, pre-trained and absorbed—with the general-purpose savoir faire that Homo Sapiens consider “common sense”. Alas, the result is an artificial idiot savant that can consistently excel at well-bounded tasks but can get things very wrong when faced with unexpected input. Bulaga! Like Van Gogh’s Starry Night, it only illuminates broadly—and pretty much like yours truly, parating na-Cailao—but struggles to pinpoint accurately… and like what Antoine de Saint-Exupery’s Little Prince said, “It is with the heart that one can see rightly; what is essential is invisible to the eye.”