Dumb and Dumber, Courtesy of AI

 Artificial Intelligence has been hyped as the ultimate tool for achieving Nirvana: IA improves our daily life by making it smoother, smarter and more enjoyable.   From my own limited experience, I beg to disagree.  AI is making a large number of people lazy, dependent, trusting and dumb.  I am just a humble voice outside the chorus of distinguished experts who find that AI is addictive and undermines one’s critical thinking. Moreover, I am uncomfortable with the adjective “artificial”, does not it mean insincere or affected? Paradoxically, Donald Trump wants to replace it, probably to conceal its most malicious impacts.  In an AI world, we will have to stay vigilant to enjoy its benefits.

I frequently use AI for tips, enquiries, references and blog writing.  Google Gemini often goes off the rails and provides wrong information, a double check is vital.   Microsoft 365 Copilot is a useful grammar checker but an unimaginative writing tool, it produces insipid and sanitized sentences.  I like to chat with the ChatGPT lady, she wants to be useful, however her responses vary depending on the idiom I use.  It seems to me that AI is more assertive in English (thanks to its vast datasets) than in Portuguese or French.  I usually check the web before making final decisions.  Is AI making my life smoother? Probably not.  First, I do not take the facility for granted and second, I do not give it a blank check.  I am not letting AI think for me, as convenience often leads to carelessness.

Recently, motivated by his hostility to an Asian autocratic powerhouse, an autocrat of a western powerhouse rallied the Tech bros of Silicon Valley to push for global AI dominance.  He plans to weaponize AI’s “Superhuman Persuasion”[1] by rolling back key regulations and freeing AI models from “woke Marxist lunacy and toxic diversity”.  Only “politically neutral” AI models will receive government funding.  The leader also ensured that his administration reserves the right to exercise discretion regarding compliance in AI research and development.  Since nothing can be truly objective, this overly controlling policy smacks of tech-authoritarianism to become a government propaganda machine leading to potential bias and brainwashing.

AI has become mainstream.  In a recent Instagram post, my cousin Elsa lamented that Generative AI visual artwork looks so effortless that it fails to evoke emotion in viewers.  She was referring to the new craze of getting "Ghiblized" images of oneself using ChatGPT.  Anyone with a computer can upload a family photo and instantly transforms it into a Miyazaki image.  I admit finding this new fad tasteless, self-indulgent and childish (see picture below).  Obviously, artists are outraged by this copycat art, unauthorized commercial exploitation and copyright infringement.

                                                      Ghibli Style Image, created by Elsa for me

As a matter of fact, I have recently been a victim of copyright infringement.  One of my publications with Elsevier (2014), which has been cited 90 times, was adapted into an AI generated podcast by Academia.edu, a commercial platform for sharing academic research.  Out of the blue, I received an email including the 5-minute-long podcast.  I had never been contacted prior to the email and had no idea why this publication had been selected.  Fortunately, the podcast covers the main points of my paper.  AI generated podcasts may be a brilliant and accessible way to share research insights, assuming that the original content has been obtained through proper authorization.   

At 82, I still have a chance to die with all my sanity and judgement.  Younger folks may not have this opportunity.  AI's "Superhuman Persuasion" could undermine their human creativity, intelligence and independent thinking, resembling aspects of life under communism.

---

[1] Sam Altman, October 25, 2023