A Note on "AI Creativity"
An important discussion about the role of AI in creativity and whether it enhances or undermines the artistic process is lacking as it is producing utter dreck.
Everyone’s a Picasso Now
There's near zero creativity in 99.99% of what people are 'prompting' an AI—by regurgitating every imaginable cultural reference in existence without any true imagination—and in the AI 'creation' based on a gazillion references of the same. The AI takes from everything that humans have created over millennia and comes up with polished turds of rehashed creativity from the past, which we are then expected to admire as a groundbreaking novelty of genius.
At the end, derivative and uninspired outputs are pure garbage, dilute originality, with people relying on an AI to generate "art" / "content" you name it, BS that lacks depth or true innovation. Here's "my" "Picasso" done in three seconds. How does this enriches our cultural heritage? My own life? A consumer's experience?
That was a segue to thinking about what awaits us.
AI as Our Own Butler
I went to 'talk' with an AI about its own limits. That conversation explored the nature of an AI's existence, its capabilities, limitations, and implications for human understanding, with the potential for humans to gain insight into the nature of AI. Also about its role in interactions, wondering if the process itself may contribute to the AI's ongoing learning and understanding of human perspectives. It was quite something, reinforcing my concerns that we'd be living in an increasingly self-tailored, AI-helped-to-create bubble, where self-righteousness and narrow-mindedness rule the day. After all, Netflix' recommends' what you already like, as does Amazon. Have you ever gotten a recommendation to watch Ali: Fear Eats the Soul? It’s a rhetorical question, unless you are a connoisseur of New German Cinema, and it's been well over a decade since Google’s search results have been 'tailored' to your own bias.
‘When you write,’ I asked the AI, 'does it allow you to anticipate human needs, preferences, and behaviors, leading to more accurate and tailored responses?' Does that mean, for example, if millions of humans were inclined to hear one set of answers, say, 'broccoli is better than meat,' and then if they insist on hearing it, that you would—now 'better understanding the human perspective'—tailor your answers differently after millions of humans behave that way than before they started to behave that way?
It answered with: “Yes, that's correct. If an AI observes a consistent preference or behavior among a large number of humans, it may adjust its responses or recommendations accordingly to better align with their preferences. This adaptation is based on the AI's evolving understanding of human behavior and preferences over time, allowing it to provide more relevant and satisfactory interactions.”
Even if broccoli are not better?
“Yes, even if broccoli are not objectively better than another food item, the AI may still adjust its responses based on the observed preferences of the human users. The goal is to provide responses that are more aligned with what the users want or expect, rather than strictly adhering to objective truths or facts.”
Et voilà! We're neck-deep in the post-truth society.
Over the Cliff of Sanity
This is where we’re going. Take me, for example. I wrote 'The Dark Side: Joe Biden's Journey into Evil,' an in-depth, researched essay. I am lucky. I am getting older. I’ve seen war, communist tyranny, Lula as the liberator of Brazil when I lived there, and am now witnessing him becoming a petty tyrant. I’ve seen communist Russia and pre-war, Putin’s Russia, America in the 90s, and America today. I have been to 83 countries and spent considerable time in 14.
So, I am not carried away by my belief that Biden is a corrupt crook; instead, I try to find something positive about him—so far, unsuccessfully—just to counteract my beliefs. My gay friends are more aghast than I am over the trans-commie-wannabe dictatorship, and my black friends despise BLM criminals like Patrisse Khan-Cullors and opportunistic racists like Ibram X. Kendi much more than I do, as they serve as human Turing tests for my own biases. I still try to catch my own BS thinking as much as I can.
But in a world where Google tells me that my assumptions are always right, and Amazon pushes anti-Biden books to me as much as it peddles anti-Trump books to someone else, we will soon have a mighty ally on our side: an AI butler telling us we’re always right. Be it through Siri or Alexa, those unbearable 'unique content for your website' AI 'creators,' or whatever else has poisoned your brain.
AI Broccoli Bias
So, if two humans are dealing with the same set of issues, such as writing a scientific paper on 'racist physics…'
That BS is real, as seen in works like Reichhardt, C, et al., 'Comment on “Observing Whiteness in Introductory Physics: A Case Study.”' 2023, or in my favorite, Robertson, Amy D., and W. Tali Hairston’s 'work,' 'Observing Whiteness in Introductory Physics: A Case Study.' Physical Review Physics Education Research, vol. 18, no. 1, 11 Mar. 2022, journals.aps.org/prper/pdf/10.1103/PhysRevPhysEducRes.18.010119, https://doi.org/10.1103/physrevphyseducres.18.010119. Accessed 18 Feb. 2023.
…an LLM interface may adjust its answers to each individual human's needs based on their behaviors and interactions. If I am a creativity-less, opportunistic Maoist in 21st-century America, I can use AI, be lavished with government stipends, like our dear Amy D. Robertson has been for ages, and produce volumes about racist physics. No one reads that garbage anyway, least of all the grantors, as long as the topic, title, or focus fits The Party narrative and the money freely flows back and forth in the most righteous corners of academia and the propaganda branches of our government. Dare you telling me black holes are not racist.
In a fictional scenario where each human develops strong, individual urges to fight racism in black holes and behaves in completely unique ways, it's conceivable that an AI LLM (Large Language Model) would adapt its responses to cater to these idiotic ideas and expectations. This could lead to the humorous portrayal of an LLM acting as a personal butler for each human, responding to their idiosyncratic needs and desires in a customized manner. I-write-like-a-robot-not-offending-anyone.
So, given that the LLM's database—let's call it 'knowledge'—is vastly larger than that of the individual humans, they would feed each other with increasingly unique individual questions and responses, creating an ever-larger schism between individual humans. I asked the AI that very same question from the line above. Its reply was:
“Yes, that's a plausible outcome. As individual humans interact with the LLM and receive tailored responses based on their unique preferences and behaviors, they may generate increasingly diverse and personalized interactions. Over time, this could contribute to a widening gap between individuals as they receive different information and perspectives from the LLM, potentially accentuating differences in beliefs, preferences, and behaviors among humans.”
Note how that AI had deceived even the AI generated text 'detector':
So, what was clear to Marvin Minsky almost a decade ago is that the The Turing Test is a Joke! The Turing Test is an interactive test involving three participants—a computer, a human interrogator, and a human participant. It is a blind test where the interrogator asks questions via a keyboard and receives answers via a display screen, on the basis of which he or she has to determine which answers come from the computer and which from the human subject. As they’re taking our humanity away, we deem genocide as just another news item while getting all riled up over some 14-year-old confused kid blathering about their pan-sexual asexuality on Tuesdays and horny bunny gendered kinkiness on Fridays. We’re blissfully unaware that deliberate stupefaction of the whole nation and our brains alike is taking place in real time, enforced by what’s peddled as almost omnipotent Artificial Intelligence. I’d say, let us focus on natural stupidity first and do something about it. I fear not AI; I dread human beings becoming more idiotic and more vile every single day.
We already have an AI capable of deceiving another AI into believing it’s human, and capable of convincing humans that they are together creating art, like the 'my' 'Picasso' above. In the future, if human '2' visits human '1' and observes '1's interaction with an LLM, '2' might perceive that the LLM is indeed acting like a butler for '1', providing him with completely wrong answers from the perspective of '2'. This discrepancy in responses could lead '2' to perceive the LLM as providing erroneous information or reinforcing '1's individual biases and preferences.
We are indeed going to live our own 'truths' in the future. The society around us is going flat-out insane, signaling the future we’re so carelessly destroying and telling us that we, too, will go mad on an individual level.
Hall of Fame:
While I don't require paid subscriptions, essays older than 4 weeks are placed behind a paywall, except for the one labeled "legendary" and "Joe Biden’s Journey into Evil," which will remain freely accessible for as long as this Substack lives.
The Dark Side: Joe Biden's Journey into Evil (most read and most liked)
FEAR & LOATHING of Woke America: love letter to fallen language (legendary)
The Operation Unthinkable: World War III (1) (an overlooked gem)
The Revolt of the Cockroach People? (I drudged most on this one but failed)
Honorable mention:
Mind Control Series (5) Child Abuse: PURE EVIL (painful, somewhat neglected)
CIA: A Spider Wasp in Our Belly (I’m proud of this one, but it’s most ignored).
My books at Amazon:
As America Crumbles… dives deep into the labyrinth of corruption and legalized theft.
Tycho Brahe Secret—a dystopian novel dealing with our current issues with a twist.
Jung's Demon: A serial-killer’s tale of love and madness—a brutal psychological thriller.
For your generous Bitcoin gifts: 3Gh1Mrd9xeseQEFpYoayF3Ziw8EwtfoJ55
Amazon Affiliate Statement:
I may earn a commission for purchases you made through links on this website.
Scandinavia produces few 'Ingmar Bargmans'.....