The Year Of Hallucinating Artificial Intelligence And Arrogant Human Ingenuity

The evolution of artificial intelligence (AI) over the past eighty years has had countless transformational moments, but 2023 may stand out as pivotal when AI took center stage in the realm of digital technology. AI has boldly characterized both the merits and pitfalls of technology, creating a range of feelings and attitudes. The pioneers of the prevailing AI model, laureates of the 2018 Turing Award for instituting fundamental advancements that established deep neural networks as indispensable in computing, articulated in 2023 their caution, admiration and modesty concerning their creation.

Caution: Geoffrey Hinton, dubbed "the father of AI", warned of potential implications of digital intelligence superseding human intelligence due to the capacity of numerous neural networks to learn simultaneously then transmit the knowledge.

Admiration: Yoshua Bengio, another pioneer of AI, expressed his astonishment at the rapid advancement of AI, contemplating if only some unknown elements were addressed, machines could potentially surpass human intellect.

Modesty: AI pioneer, Yann LeCun highlighted that while advances have been made, there's still considerable undiscovered ground that's necessary to bring machine learning on par with human or animal learning.

In Alan Turing's 1950 paper "Computing Machinery and Intelligence", he proposed evaluating machine intelligence by assessing if machines are capable of emulating human-like activities. This is now known as the "Turing Test". However, successfully mimicking human dialogue doesn't necessarily imply that a machine thinks like a human.

Five years following Turing's publication, the term "artificial intelligence" was coined by John McCarthy, ever since, the field of AI has been striving to create convincing simulations of human-like capabilities. Nonetheless, thus far, authentic human intelligence remains unattainable for machines, despite continuous advancements.

The widespread belief in 2023 is that we're entering a new digital era, akin to the advent of the internet in 1993 and smartphones in 2007. Moreover, 2023 was the year Ilya Sutskever, OpenAI's chief scientist, made the bold proclamation of the imminent arrival of artificial general intelligence, or even superintelligence.

OpenAI, in 2023, suggested that AI, being highly impactful, might precipitate human disempowerment or even extinction if unchecked. This prediction was associated with the philosophical notion that humans are analogous to gods, with the capability of making radical changes using technology.

AI has certainly demonstrated its potential, some have become convinced that AI needs only to resolve the issue of "hallucinations", or untrue or false responses from AI models, to reach human-level smartness. The term "hallucinations" was initially used in image recognition context but it has extended to all AI context implying discrepancies in AI outputs.

While celebrating human ingenuity in advancing AI, sophistication of technology doesn't imply superhuman capabilities. As the new year unfolds, it's hoped that humility will supplant grandiose notions and claims, encouraging a balanced and realistic perspective on the future of AI.