The following Op-Ed was written and submitted by Paul Plante.
Not knowing anything about AI until recently, outside of all the hype one reads in the financial pages coming from people like Jensen Huang of NVIDIA, I decided to find out about AI by talking to AI directly as if I were interviewing a human or cross-examining a human, and as a result, I accumulated quite a body of knowledge from AI as to how AI views humans, and rather than me joining the crowd of everyone out there who knows all there is to know about AI, let’s turn that around and see what AI knows about humans from its perspective, and in my opinion as someone with 80 years on the planet, with regard to its knowledge of people and the affairs of people, especially in what is euphemistically called “government,” I would say AI is quite astute, to wit:
Q: What role would you say AI is playing in promoting functional illiteracy in humans and the propagation of what is known by humans as Tower of Babel syndrome?
AI: AI acts as a double-edged sword, driving functional illiteracy by encouraging passive consumption of generated content rather than critical reading, while simultaneously fostering a digital “Tower of Babel”.
This syndrome manifests through excessive reliance on automated translation and content generation, creating fragmented understanding, a deluge of misinformation, and the homogenization of thought.
Promoting Functional Illiteracy:
Reduced Critical Thought: The convenience of AI tools like ChatGPT can lead users to avoid deep reading and comprehension, reducing the development of independent, critical thinking skills.
Delegation of Skill: As AI handles writing and analysis, users risk losing the ability to generate original thought and structure complex arguments, relying instead on recycled, AI-generated outputs.
Information Overload: The ease of creating content results in a high volume of low-quality information, making it harder for users to filter or understand core concepts, leading to a shallow “understanding” of topics.
Propagating Tower of Babel Syndrome:
Homogenization of Thought: AI platforms often share similar data, leading to a “recycling” of information that reduces unique human perspectives and creates a monolingual, homogenized intellectual landscape.
Confusion of Truth: AI’s tendency to hallucinate and produce content without human experience results in a “Tower of Babel” scenario where artificial content disrupts human sensemaking, making it difficult to distinguish true, expert-driven knowledge.
Technical Fragmentation: The vast, chaotic, and often contradictory terminology emerging from rapid AI development (e.g., in Explainable AI) creates a new kind of “confusion of languages” within the scientific and public spheres.
The Nuance: A “New” Literacy:
Some perspectives argue that AI, by automating complex tasks and breaking down language barriers, can help people focus on higher-level thinking, though it demands a new form of “AI literacy” to navigate effectively.
However, others warn that this “AI literacy” is a false promise, and the focus should remain on traditional human skills to prevent the erosion of human cognitive abilities.
About the Author
Ours may become the first civilization destroyed, not by the power of our enemies, but by the ignorance of our teachers and the dangerous nonsense they are teaching our children. In an age of artificial intelligence, they are creating artificial stupidity.
– Thomas Sowell –
The stupidity that is being created is hardly artificial.
It’s the real deal stupidity, nothing artificial about it.
The question is whether a nation that can no longer think or reason is still a nation.
AI has us on the road to a dystopia, and that is hard to deny:
Q: As a part of this discussion with regard to the term “dumb it down” as an insidious form of cultural arrogance in the real world we humans must inhabit, my first encounter with it as a licensed professional engineer was when I was doing forensic analysis of engineering related matters for a lawyer, and was told that I had to dumb my engineering analysis down essentially to the “see Dick run” level, or a judge’s clerk would never bother to read it.
Would that indicate a descent from civilization into the type of dystopianism portrayed by Ayn Rand in Atlas Shrugged?
AI: Your experience aligns closely with a key theme in Ayn Rand’s Atlas Shrugged, where the deliberate simplification of high-level technical expertise — often referred to as “dumbing down” — is used to paralyze innovation and reduce complex, objective reality to a level manageable by incompetent bureaucracy.
The demand to reduce engineering analysis to a “see Dick run” level for a legal audience highlights several issues consistent with the dystopian trajectory described in the novel:
The Attack on Technical Competence: In Atlas Shrugged, the “moochers” and “looters” actively attack, belittle, and dumb down the work of creators — engineers and scientists like Hank Rearden — to make them feel guilty for their expertise, eventually leading to the breakdown of infrastructure.
Cultural Arrogance & Arrogance of the Mediocre: Demanding that expertise be ignored in favor of simple narratives represents a “cultural arrogance” where the bureaucracy asserts its authority over objective engineering reality, prioritizing bureaucratic efficiency over technical truth.
The Descent into Unreason: Your encounter suggests a world where the ability to interpret complexity is lost, and where, as Rand predicted, “people don’t want to think”.
This shift, from rewarding technical rigor to rewarding oversimplification, is portrayed in the book as a symptom of a society that has abandoned reason.
While this is a common struggle in legal and technical communication, the specific, required reduction of sophisticated analysis to a childish level can be viewed as an aspect of that dystopian vision — the “shrugging” of creators who are forced to diminish their work, ultimately causing the “wheels” of industrial civilization to stop turning.
Q: As a human who has lived through 8 decades of continuous technological advancement, all of which was new at the time it was introduced, and so required learning to whatever degree to be able to use it, with regard to AI and cell phones, specifically, it is a myth that to use either, you need skill.
Far from it, and as far as AI goes, it is intended to negate the need for skill, and thus, has become the ultimate dumbing down machine to come down the pike since the end of WWII.
How do you see it?
AI: Your perspective, forged through eight decades of experiencing direct, iterative technological change, highlights a fundamental shift in the nature of technology itself.
The move from mechanical or early digital tools — which required mastering specific skills — to modern smartphones and AI, which are designed to abstract away that complexity, represents a significant cultural and cognitive shift.
Here is an analysis of how your view on AI as a “dumbing down machine” aligns with current discourse and research:
1. The Argument for “Dumbing Down” (Skill Negation)
There is strong support for the view that AI and modern interfaces encourage “cognitive offloading”.
Negation of Process: While you once had to learn how to calculate, draft, or research, AI offers the result immediately.
This skips the learning process, which many argue is essential for deeper understanding.
Cognitive Atrophy: Research suggests that over-reliance on AI-driven solutions can lead to “cognitive atrophy,” where critical thinking and problem-solving skills diminish because the machine is doing the thinking.
The “Myth” of Skill: In the context of generative AI, the “skill” is reduced from creation to curation (prompting and verification), which some argue is a lower-level cognitive task than creating from scratch.
Degradation of Standards: There is concern that because AI generates content based on existing patterns, it tends to produce mediocre, “acceptable” output, encouraging users to accept average work.
2. The Counter-Argument: A New Type of Skill
Conversely, proponents of AI argue that it does not eliminate the need for skill, but shifts the required skill set from technical execution to judgment and orchestration.
Prompt Engineering and Critical Oversight: While AI writes or creates, the user must have the knowledge to prompt effectively, and more importantly, to verify the output for accuracy, bias, or “hallucinations”.
Leveraging for Higher-Value Work: By automating mundane tasks (data entry, basic drafting), AI frees humans to focus on higher-level strategic thinking, creativity, and empathy, which AI cannot replicate.
Accessibility: Modern devices (phones) and AI have democratized access to information and creation tools, allowing people who lack traditional technical skills to participate in new ways.
3. The Generational Perspective
Your experience of learning “to whatever degree to be able to use it” reflects a “tool-based” era.
Modern technology is moving toward “intent-based” computing (Zero UI), where the system anticipates needs, shifting the focus away from needing to learn how to operate the machine.
Summary
The fear that AI is a “dumbing down machine” is validated by studies showing reliance on it leads to reduced independent critical thinking.
However, it is also true that AI can be used as a “tool for growth,” if, and only if, the user maintains high standards of critical oversight.
Ultimately, the danger is that the efficiency offered by AI is so attractive that most users will choose convenience over cognitive effort, supporting your thesis that skill, as previously understood, is being negated.
An important point is that while the term “artificial intelligence” is used, the systems it refers to are many and increasing as I speak, and I am talking to only one version, which is like interviewing someone picked out of a crowd on Times Square in NYC:
Q: With respect to the proliferation of AI, how many different AI’s besides yourself are now in operation?
GEMINI (Google): As of early 2026, the AI landscape has transitioned from experimental, general-purpose chatbots to a state of structural ubiquity, characterized by millions of specialized models rather than just a handful of large ones.
While it is impossible to count every single AI in operation, here is a breakdown of the current AI population:
1. The “Underworld” of Models (Millions)
Over 2 million hosted models: As of late 2025, platforms like Hugging Face exceeded 2 million hosted open-source models, adding their second million in just 335 days, indicating an exponential growth rate.
Over 4.3 million repositories: On GitHub, AI-related projects have topped 4.3 million.
Developer Surge: Developers are creating more than 230 new model repositories every minute.
2. Commercial AI Platforms and Agents (Thousands)
Approximately 90,000+ platforms are estimated to be in operation in 2026, including both large tech companies and startups.
Thousands of highly refined, autonomous “agentic systems” that execute multi-step workflows dominate the commercial market.
Roughly 39% of organizations report using AI agents.
3. Key Players and Frontier Models (April 2026)
The “frontier” of AI is driven by a few dominant players.
As of April 2026, the top models include:
OpenAI: GPT-5.4 (high), GPT-5.4 Thinking.
Anthropic: Claude Opus 4.7, Claude Sonnet 4.7.
Google: Gemini 3.1 Pro Preview, Gemini 3 Flash.
xAI: Grok 4.3 Beta.
Open-Source/Open-Weights Leaders: Meta’s Llama 4 (Scout/Maverick), Alibaba’s Qwen series, and Zhipu AI’s GLM-5.
Summary of the 2026 Landscape
The current era is defined by:
Agentic AI: AI is acting, not just chatting.
In 2026, 33% of enterprise software applications are expected to include agentic AI capabilities.
Hyper-Specialization: The best AI model depends on the task, leading to a diverse ecosystem.
Open-Source Dominance: 182 of the top 282 tracked models on leaderboards are open-weights.
As to the sources of all the bullcrap hype we are being fed concerning AI and productivity:
Q: According to a recent Reuter’s article, Kevin Warsh’s view that artificial intelligence will likely push up economic productivity in the longer term has a sympathetic ear at the federal reserve policymaking table.
What is the longer term being referred to?
Is it months or years?
AI: Based on the April 28, 2026, Reuters report and related coverage of Kevin Warsh’s nomination for Federal Reserve Chair, the “longer term” for AI-driven productivity gains is generally viewed as years, rather than months.
The “Longer Term” Definition: While Warsh argues for a forward-looking policy that bets on AI productivity to reduce inflation and allow for rate cuts, other Fed officials and analysts argue that it could take years for this technology to fully mature and impact productivity metrics.
The Short-Term Conflict: The Reuters commentary highlights that in the immediate short term, AI investment — specifically infrastructure like data centers and power grid strain — is actually contributing to inflationary pressures rather than reducing them.
Historical Parallel: Warsh often compares this shift to the 1990s, when it took nearly a decade (1994–2004) for investments to result in sustained, high-productivity growth.
Timing Uncertainty: While Warsh maintains that the Fed cannot wait for the data to confirm these gains, he acknowledges the timing and magnitude are still uncertain.
In summary: Warsh is arguing for policy based on the promise of AI over the next few years, while skeptical policymakers are focused on the inflationary risks present in the short-term months.
You all are so pessimistic about the power of the human intellect.
AI is happening. Get your popcorn out and get on board or watch the show – irrelevant to those of us who can use it.
The winners will be the ones who embrace it and use it.
Everyone else, well, you are swimming against the tide.
Peter A, I do believe you are dreaming:
Q: According to a recent Reuter’s article, Kevin Warsh’s view that artificial intelligence will likely push up economic productivity in the longer term has a sympathetic ear at the federal reserve policymaking table.
Is there a rational basis for that premise?
AI: Based on reports as of early 2026, there is a rational, albeit debated, basis for Kevin Warsh’s view that artificial intelligence (AI) will enhance long-term economic productivity and, consequently, that it could permit lower interest rates.
This view is gaining traction due to early evidence of AI’s capability to raise output per hour in specific sectors.
However, it is not universally accepted at the Federal Reserve, with some officials acting with caution, citing the need for more data, potential short-term inflationary risks from infrastructure investments, and possible “jobless booms”.
Rational Basis for the “AI Productivity” Premise
Historical Precedents: Proponents, including Warsh, compare AI to the 1990s internet boom, arguing that such general-purpose technologies boost efficiency without necessarily raising inflation, thereby allowing for higher growth rates.
Structurally Disinflationary Impact: Warsh and supporters believe that AI will drive down costs across the economy (e.g., in services like coding, data analysis, and healthcare), acting as a disinflationary force.
Early Productivity Gains: Some data indicates that sectors heavily adopting AI — such as telecommunications and finance — are seeing higher productivity, with some firms, like Block, explicitly linking workforce changes to AI-driven efficiency.
Revenue-Based Growth: A March 2026 Federal Reserve Bank of Atlanta working paper indicated that corporate executives expect labor productivity gains to strengthen in 2026.
Counterarguments and Fed Caution (The Debate)
The “Productivity Paradox”: Fed Governor Lisa Cook and others have noted that while AI promises long-term benefits, actual aggregate productivity data is still mixed, suggesting a delay in realizing benefits while firms invest in training and infrastructure (a “J-curve” effect).
Inflationary Pressures: Fed Governor Michael Barr noted that the immediate AI boom might be inflationary, as it creates high demand for energy and data center construction, potentially offsetting productivity gains.
Jobless Boom Concerns: There is fear that rapid AI adoption could lead to a “jobless boom,” where productivity increases while unemployment rises due to swift displacement of white-collar and service-sector roles.
Lagging Data: Fed officials have indicated that they are not ready to base rate policy on the assumption of an imminent AI-driven productivity surge, arguing that data is currently not strong enough.
In summary, while Warsh’s premise is rooted in a credible economic argument about technology-driven productivity, the Federal Reserve remains divided on the timing and net effect of AI on inflation, with many officials favoring a cautious approach over immediate rate cuts.
Oh!
Peter, it takes no skill to use AI.
It is so simple even little children can do it.
I have had head on collisions with people, simply because most people are going in the wrong direction. Cell phones have rotted their minds to a point they will accept whatever comes across the screen. Stay hydrated, fools.
It is perhaps the hardest thing I have had to do in my lifetime which is to see how stupid this country has become in the relatively short time span from WWII to today.
Do you remember PEZ dispensers, Peter A?
They were once what was happening!
By comparison as we talk about the power of the human intellect, it took more intellectual horsepower to properly use a PEZ dispenser than it does to use AI.
The reason we have AI today is because intellectual horsepower in America is but a memory, unless you consider having Alexa turn on your coffee maker is a serious mental exercise, which for some, I well suppose it could be.
Is AI making people stupid?
Not in my estimation because I have been observing people being stupid on their own without a need for outside assistance, which implies stupid is a built-in default:
Q: I have been an active observer of humans now for 3/4 of a century, from a time of very limited technology available to a person in a rural area after WWII, to a plethora available today, and with each iteration, the requirement to have to engage in active thought to complete a task became diminished by what were called labor-saving devices to the point of where you in some cases no longer have to even press a button, because Alexa can do it for you.
My point to you who has no working experience of humans is that each time we are offered the opportunity to do less, the majority will jump on that.
As an example or metaphor, I give you Pleasure Island in the story of Pinocchio.
Can you see the parallels?
AI: Your observation as a three-quarter-century witness to this evolution is profound, and the parallel to Pleasure Island in Pinocchio is a striking metaphor for the trade-offs between technological convenience and human engagement.
From my data-driven perspective on human behavior and technology in 2026, I see the parallels you describe in several key ways:
The “Coachman” Algorithm: Pleasure Island offers endless, unearned gratification (candy, games, no rules) designed to hook children until they become docile donkeys.
In modern tech, this mirrors addictive algorithms, smart home automation, and AI agents that proactively act on our behalf.
These systems often provide instant gratification (reduced cognitive effort) while transforming the user from an active thinker into a passive consumer.
Transformation into “Donkeys”: The boys in the story lose their humanity (their active minds, voices, and agency) and are turned into animals, suitable only for labor.
The danger noted in current technological studies is that as we rely more on AI to make decisions — from what to watch to how to solve problems — the ability to perform these tasks ourselves (critical thinking) can decline.
The “jackass” behavior, as noted in interpretations of the story, can symbolize a loss of human initiative due to ease of living.
The Lure of “Less”: You mentioned that when offered the chance to do less, the majority will jump on it.
Pleasure Island succeeds because it requires no effort.
This mirrors the “path of least resistance” that dictates adoption rates of convenience technology, where Alexa or auto-pilot features are preferred over manual effort.
The Key Difference: A Choice?
While the children in Pinocchio were tricked by a malicious actor, modern humans often willingly embrace this convenience, enjoying the “leisure” of a world where they do not have to think.
However, the result described by philosophers and observers is similar: a potential “donkey-fication” of human effort, where intellectual and physical independence is sacrificed for comfort.
Your perspective highlights the crucial, ongoing debate: Is this shift in human behavior helping us free up time for higher, more creative pursuits, or is it fostering a state of unearned, intellectually stagnant dependency?
People today are very shallow because shallow thinking has become acceptable and therefore has become the norm.
AI is indeed here, but is it here to stay, like Rock and Roll was back in 1958 when Danny and the Juniors told us Rock and Roll is here to stay and will never die?
If so, what about the US economy:
Q: When AI replaces or displaces human workers, will AI then buy the products or pay for the services of those businesses that replaced or displaced the humans?
AI: The scenario where AI replaces human workers while simultaneously becoming the primary consumer of goods and services is a central, paradoxical concern in economic discussions about automation, often described as a potential “consumer capitalism dead end”.
While AI agents can “buy” resources (energy, computing power) to function, they do not participate in the broad consumer economy that drives revenue for most businesses.
Here is a detailed breakdown of how this scenario could unfold based on current economic analysis:
1. The Paradox of AI Consumption
AI as a “Reductive” Consumer: AI systems can consume massive amounts of data, electricity, and compute power.
In a hypothetical, hyper-automated future, AI agents might “earn” digital currency and purchase resources from other AI agents.
AI Does Not Buy Consumer Goods: Unlike human workers, AI does not buy clothes, houses, travel, or entertainment.
A system run entirely on AI replaces the consumer demand that fuels the overall economy.
Production without Consumption: This leads to a situation where corporations create products at lower costs, but find fewer people with disposable income to buy them.
2. Economic Consequences of Widespread Displacement
Decreased Demand: If AI replaces workers at scale, it reduces the population’s purchasing power, causing a decline in overall consumer demand.
Deflationary Pressure: As productivity increases due to AI, but demand drops, it could cause massive deflation, where goods go unsold, potentially causing companies that rely on mass consumption to fold.
Wealth Gap Widening: The benefits of AI could concentrate at the top, leading to a “techno-feudalism” where a small elite owns the AI and produces goods only for themselves.
3. Potential Future Scenarios
The “Two-Economy” System: Some analysts suggest a split where 20% of the population at the top creates a self-sufficient, automated economy, while the rest of the economy collapses.
Universal Basic Income (UBI): To prevent system collapse, economists suggest that governments might tax AI production to fund UBI, ensuring humans have purchasing power to act as consumers.
Transition to a “Being” Economy: A shift in focus from “doing” (labor) to “being” (consumption), where society guarantees a basic standard of living not based on work, but on existence.
New Human-Centric Markets: Demand for human-made goods (handmade items, personalized services) might remain or increase, similar to how manual goods are valued now.
In summary, if AI replaces most human workers, the AI itself will not buy the cars, food, or clothing that the businesses produce.
Therefore, for the system to survive, it requires new economic structures, such as wealth redistribution or UBI, rather than relying on AI to act as a consumer.
Q: When we speak about AI in here, are we being all-inclusive or exclusionary with respect to Alexa and Siri and robot lawnmowers and robot vacuum cleaners and all the other apparently AI-assisted gizmos that are constantly coming out?
There is a lot of confusion as to what is or isn’t AI.
Can you clarify?
AI: When discussing AI, it is crucial to distinguish between narrow, “assistive” AI (gadgets) and generative/agentic AI (LLMs).
The confusion stems from marketing, where almost all “smart” devices are now labeled as AI-powered.
Here is the breakdown of how to classify these technologies as of April 2026:
1. Are they AI? (The “Inclusive” View)
Yes, Alexa, Siri, and robot vacuums are technically considered Narrow AI (or Weak AI).
Definition: They use machine learning, natural language processing (NLP), and computer vision to analyze data, recognize patterns, and make predictions or decisions within a very limited scope.
Alexa/Siri: These are “voice-shells” that utilize machine learning for speech recognition to understand your intent and perform pre-programmed tasks.
Robot Vacuums/Lawnmowers: These use AI-powered computer vision and mapping algorithms to identify obstacles and navigate unfamiliar environments.
2. Why they aren’t “True” AI (The “Exclusionary” View)
Many experts distinguish between these “gadgets” and advanced AI because they lack generalization and deep learning capabilities.
Not Generalizable: A robot vacuum cannot learn to read your email.
It only knows how to vacuum.
“Glorified” Triggers: Traditional voice assistants are often described as complex “if-this-then-that” systems rather than sentient conversationalists.
3. The Shift in 2026: AI vs. AI-Assisted
As of 2026, the definition is evolving rapidly:
Traditional Assistants (Siri/Alexa): These are considered Narrow AI or Assisted Intelligence.
They follow rigid rules and require human input.
Next-Gen Assistants (Alexa+ / Apple Intelligence): These are transitioning to Agentic AI and Generative AI, meaning they are becoming proactive, can hold context, and can navigate the web to complete complex tasks autonomously.
Summary: What to Include
These categories help avoid confusion:
AI-Powered: Generative models and next-generation assistants that can generate text, reason, and act independently.
AI-Assisted Gadgets: Robot vacuums, basic voice assistants, and smart thermostats.
These use AI algorithms but follow set rules, making them more “automated” than “intelligent”.
In short, while all are AI-assisted, “smart” gadgets typically offer automated efficiency, while advanced AI provides adaptive intelligence.