The Only Three Plays That Might Spur AI To Take Over

Pradeep Aradhya
5 min readMay 4, 2023

ChatGPT, GPT-4, Bard etc. have all raised the AI euphoria to fever pitch. And then there are any number of voices crying caution and foul and demon. Educators, Italy, the EU, and even Elon Musk are urging caution and slow down and even stop. They have ignited their own fear fever about everything from AI disasters to AI take over. Suddenly people want to know if AI feels emotion, what with all the AI counselling going on. Suddenly everyone is asking not whether AI will become sentient but when. Suddenly, there are discussions and debates and webinars about whether AI will take over.

Many an expert — computational, psychological, and philosophical has opined about AI taking over. Theories and ideas abound — ranging from war gone out of control to when AI swarming is used prematurely to biochemical armageddon to financial market mayhem. Apparently and thankfully AI has already killed the metaverse. None, however explore the exactly how or why scenarios for this much implied and conjectured AI domination of humanity.

There are three plays that are prerequisites to precipitate an AI take over. Thankfully, they are disjoint and easily enough controlled for us to prevent it. Unfortunately, they are just as easy to forfeit in ignorance.

Nearly 30 years ago, I asked an erstwhile academic who choose to enlighten a bunch of us young guns about AI whether it would take over. In the last 5 years, I have had everything from former film associates try to characterize my movie’s AI to act emotional and even coy to professionals in many spaces who claim AI cannot do their jobs. In particular, this past year has been a bunch of fierce back and forth about AI and creativity.

The bottom line though is that AI can only recombine its training data set … and mandates … even if it did unsupervised deep learning and was atleast limited memory. It might reveal a heretofore untraversed combination or data pattern prediction but it will stay remain within the meta boundaries of its training data set and mandates. It will neither be able to credibly interpret nor answer a question outside that training set. It definitely will not contravene its mandates. And while the idea of “identity” and even schizophrenia might be trained into it, “self awareness” still eludes most humans. How then…

--

--

Pradeep Aradhya

Exploring boundaries on culture, business strategy, and technology. Film maker, Kidlit Author, Technologist, Philanthropist, Investor, CEO