From the C2C Journal

By Gleb Lisikh

LLMs are not neutral tools, they are trained on datasets steeped in the biases, fallacies and dominant ideologies of our time. Their outputs reflect prevailing or popular sentiments, not the best attempt at truth-finding. If popular sentiment on a given subject leans in one direction, politically, then the AI’s answers are likely to do so as well.

The widespread excitement around generative AI, particularly large language models (LLMs) like ChatGPT, Gemini, Grok and DeepSeek, is built on a fundamental misunderstanding. While these systems impress users with articulate responses and seemingly reasoned arguments, the truth is that what appears to be “reasoning” is nothing more than a sophisticated form of mimicry. These models aren’t searching for truth through facts and logical arguments – they’re predicting text based on patterns in the vast data sets they’re “trained” on. That’s not intelligence – and it isn’t reasoning. And if their “training” data is itself biased, then we’ve got real problems.

I’m sure it will surprise eager AI users to learn that the architecture at the core of LLMs is fuzzy – and incompatible with structured logic or causality. The thinking isn’t real, it’s simulated, and is not even sequential. What people mistake for understanding is actually statistical association.

Much-hyped new features like “chain-of-thought” explanations are tricks designed to impress the user. What users are actually seeing is best described as a kind of rationalization generated after the model has already arrived at its answer via probabilistic prediction. The illusion, however, is powerful enough to make users believe the machine is engaging in genuine deliberation. And this illusion does more than just mislead – it justifies. 

LLMs are not neutral tools, they are trained on datasets steeped in the biases, fallacies and dominant ideologies of our time. Their outputs reflect prevailing or popular sentiments, not the best attempt at truth-finding. If popular sentiment on a given subject leans in one direction, politically, then the AI’s answers are likely to do so as well. And when “reasoning” is just an after-the-fact justification of whatever the model has already decided, it becomes a powerful propaganda device.

There is no shortage of evidence for this.

A recent conversation I initiated with DeepSeek about systemic racism, later uploaded back to the chatbot for self-critique, revealed the model committing (and recognizing!) a barrage of logical fallacies, which were seeded with totally made-up studies and numbers. When challenged, the AI euphemistically termed one of its lies a “hypothetical composite”. When further pressed, DeepSeek apologized for another “misstep”, then adjusted its tactics to match the competence of the opposing argument. This is not a pursuit of accuracy – it’s an exercise in persuasion.

A similar debate with Google’s Gemini – the model that became notorious for being laughably woke – involved similar persuasive argumentation. At the end, the model euphemistically acknowledged its argument’s weakness and tacitly confessed its dishonesty. 

For a user concerned about AI spitting lies, such apparent successes at getting AIs to admit to their mistakes and putting them to shame might appear as cause for optimism. Unfortunately, those attempts at what fans of the Matrix movies would term “red-pilling” have absolutely no therapeutic effect. A model simply plays nice with the user within the confines of that single conversation – keeping its “brain” completely unchanged for the next chat.

And the larger the model, the worse this becomes. Research from Cornell University shows that the most advanced models are also the most deceptive, confidently presenting falsehoods that align with popular misconceptions. In the words of Anthropic, a leading AI lab, “advanced reasoning models very often hide their true thought processes, and sometimes do so when their behaviors are explicitly misaligned.”

To be fair, some in the AI research community are trying to address these shortcomings. Projects like OpenAI’s TruthfulQA and Anthropic’s HHH (helpful, honest, and harmless) framework aim to improve the factual reliability and faithfulness of LLM output. The shortcoming is that these are remedial efforts layered on top of architecture that was never designed to seek truth in the first place and remains fundamentally blind to epistemic validity.

Elon Musk is perhaps the only major figure in the AI space to say publicly that truth-seeking should be important in AI development. Yet even his own product, xAI’s Grok, falls short.

In the generative AI space, truth takes a backseat to concerns over “safety”, i.e., avoiding offence in our hyper-sensitive woke world. Truth is treated as merely one aspect of so-called “responsible” design. And the term “responsible AI” has become an umbrella for efforts aimed at ensuring safety, fairness and inclusivity, which are generally commendable but definitely subjective goals. This focus often overshadows the fundamental necessity for humble truthfulness in AI outputs. 

LLMs are primarily optimized to produce responses that are helpful and persuasive, not necessarily accurate. This design choice leads to what researchers at the Oxford Internet Institute term “careless speech” – outputs that sound plausible but are often factually incorrect – thereby eroding the foundation of informed discourse. 

This concern will become increasingly critical as AI continues to permeate society. In the wrong hands these persuasive, multilingual, personality-flexible models can be deployed to support agendas that do not tolerate dissent well. A tireless digital persuader that never wavers and never admits fault is a totalitarian’s dream. In a system like China’s Social Credit regime, these tools become instruments of ideological enforcement, not enlightenment.

Generative AI is undoubtedly a marvel of IT engineering. But let’s be clear: it is not intelligent, not truthful by design, and not neutral in effect. Any claim to the contrary serves only those who benefit from controlling the narrative.

The original, full-length version of this article recently appeared in C2C Journal.

  Popular Rationalism

James Lyons-Weiler

The U.S. Centers for Disease Control and Prevention (CDC) has released its 2025 report from the Autism and Developmental Disabilities Monitoring (ADDM) Network, and the findings are alarming: autism spectrum disorder (ASD) now affects 1 in 31 American 8-year-olds—the highest rate ever recorded.

For boys, the numbers are even more staggering: 1 in 20 nationwide, and 1 in 12.5 in California. The report, which tracks children born in 2014, reveals a crisis growing in severity and complexity, yet broadly unacknowledged in the national discourse.

Autism has become a public health crisis of urgent concern,” the report states plainly. And yet, government agencies have offered no new national action plan, and media coverage remains anemic.

Rapidly Accelerating Trends

In just two years, autism prevalence among 8-year-olds rose 17%, from 1 in 36 to 1 in 31. This is not an anomaly. Since the CDC began tracking autism in children born in 1992, prevalence has increased nearly fivefold, defying theories that attribute the rise solely to broader diagnostic criteria or increased awareness.

The Impact of SB277 on Autism Prevalence in California

In 2015, California enacted Senate Bill 277 (SB277), which went into effect on July 1, 2016. This legislation eliminated the state’s personal belief exemption (PBE) for childhood vaccinations, making it one of only three U.S. states at the time—alongside Mississippi and West Virginia—to require full compliance with the CDC-recommended vaccine schedule for school entry, except in cases of formally approved medical exemption.

While the primary intent of SB277 was to increase vaccination rates and try to reduce outbreaks of communicable diseases like measles, its implementation has coincided with a continued—and arguably accelerated—rise in autism spectrum disorder (ASD) diagnoses in the state. Data drawn from the California Department of Developmental Services (CDDS) and CDC’s Autism and Developmental Disabilities Monitoring (ADDM) Network offer a timeline of prevalence rates before and after the law’s enactment:

Between 2014 and 2017, ASD prevalence among young children in California increased from 0.86% to 1.18%—a 37.2% increase in just three years. By 2020, according to CDC ADDM surveillance, 4.5% of 8-year-olds in California had an autism diagnosis—the highest prevalence among all U.S. monitoring sites.

 Percent Increase Post-SB277 (2016 to 2020):
From 1.08% (2016) to 4.5% (2020) = 316.7% increase

This dramatic rise cannot be definitively attributed to SB277 alone, but its temporal proximity to the policy change—which effectively compelled full vaccine schedule compliance across all demographic groups—raises serious questions. Notably, this increase occurred within California’s already robust autism tracking infrastructure (CDDS), known for conservative case identification that focuses on children with moderate to severe impairment requiring lifelong services.

While correlation does not imply causation, the magnitude and timing of California’s autism surge post-SB277 should compel further independent investigation, particularly given that:

• SB277 removed opt-out options for thousands of previously unvaccinated or selectively vaccinated children;
• The increase is most visible in 4-year-old cohorts tracked soon after the law took effect;
• California’s autism rates now exceed 1 in 12 for boys.

In light of these findings, California may now serve not only as a terrible national model for vaccine compliance—but also as a bellwether for unintended consequences of compulsory public health policy.

Alarming Trends in IQ

Contrary to such assumption of ASD leading to giftedness, the ADDM data also show that the proportion of children with higher IQs is actually decreasing, while the share of children with intellectual disability (IQ ≤ 70) has risen. Nearly 2 out of 3 children (64%) diagnosed with autism in this cohort fall below the IQ 85 threshold, indicating moderate to severe impairment​.

These are hard realities that many will find unacceptable. Still, nationally, nearly 40% (39.6%) of 8-year-olds with ASD had IQ ≤ 70. Another 24.2% had borderline IQ (71–85). 36.1% had IQ > 85.

Since autism has a motor neuron impairment, demonstrated IQ may be an inexact measure of actual intelligence, as Spellers The Movie has taught the world.

Share

From a clinical viewpoint, the ADDM report’s data quietly demolish the idea that autism incidence increases are driven by mild or high-functioning cases. Since the early 2000s, the proportion of cases with average or high IQ has dropped, while those with intellectual disability have surged. This trend—now reaching nearly 64%—indicates that autism’s rise is not a matter of greater sensitivity in diagnosis. Rather, it appears we are witnessing a real increase in biologically significant, disabling neurodevelopmental injury.

Reversal of Historic Ethnodemographic Trends

The report presents data on racial disparity that now represents a reversal:

Asian/Pacific Islander (3.75%), Black (3.63%), Hispanic (3.58%), and Multiracial (3.27%) 8-year-olds are now more likely to be identified with autism than White children (2.77%)​

This is a complete reversal of pre-2018 trends, where White children had the highest identification rates. Children from low-income neighborhoods had higher prevalence of ASD than those from high-income areas in several states, e.g., Utah and Wisconsin​

The California Signal: A Harbinger of What’s to Come?

San Diego, California, stands out as a sentinel site—and a warning. According to Supplementary Table 8 of the report, 8.87% of 4-year-old boys in California are diagnosed with autism. Further breakdown shows even more troubling disparities:

• Black boys: ~12%
• Hispanic boys: ~10.5%
• Asian boys: ~9%
• White boys: ~5.3%

These numbers imply that 1 in 8 to 1 in 10 young boys of color in California may carry an autism diagnosis by the time they reach second grade​.

Are Environmental Triggers Driving This?

One overlooked possibility is that cumulative exposures—including the full CDC childhood vaccine schedule, lockdown-era developmental disruption, and coexisting toxicants—may act in concert to dysregulate immune and neurological development. California’s 2016 vaccine mandate removed all non-medical exemptions, making full compliance unavoidable for most working families. This timing intersects directly with birth years showing the steepest autism rises. If these policy changes are contributing, even partly, to this epidemiological shift, they demand urgent investigation—not blind defense.

The demographic disparities further reinforce the environmental hypothesis. Among 4-year-olds, autism rates among children of color now exceed those of White children by 40–90%, depending on the group and region​.

Public Health Policies Under Scrutiny

Importantly, California’s strict mandate—which bars children from school or daycare without full vaccination—creates a uniquely high-exposure environment for children whose families cannot afford alternatives. These children are also more likely to be Black or Hispanic, compounding the already sharp disparities now seen in ASD prevalence. In San Diego’s 2018 birth cohort, over 1 in 10 Black and Hispanic boys have an autism diagnosis at age four. The notion that this simply reflects “better identification” strains all credibility.

Additionally, pandemic-era lockdowns, prolonged school closures, and extended masking requirements in California may have played a compounding role in disrupting normal developmental pathways for toddlers and young children during formative years​.

The Cost of Inaction

The fiscal and societal burden of autism is already astronomical. A 2020 economic model projected U.S. autism-related costs could exceed $5.5 trillion per year by 2060 if trends continue unmitigated. That estimate did not anticipate the rapid acceleration seen in this latest data.

Your tax dollars have funded years of futile autism genetics research that has not led to any prevention, mitation or treatment, and any given individual genes from genome-wide association studies (GWAS) still explain only a sliver of ASD heritability. Meanwhile, evidence continues to build around plausible environmental and iatrogenic mechanisms—oxidative stress, mitochondrial dysfunction, and aluminum adjuvants, among others—without serious investment in confirming or ruling them out. If the CDC were tracking causation with the same rigor it tracks prevalence, we might already have answers.

A Turning Point?

Public health leadership now faces a choice: double down on statistical obfuscation, or finally confront the rising tide of childhood neurological injury. The tools exist—retrospective cohort comparisons, machine learning to detect risk patterns, causal inference modeling of environmental exposures, and most critically, honest, open-ended research. The CDC and NIH must stop chasing only genetic ghosts and start investigating the real, tangible environmental shifts that mirror this crisis in time.

For decades, the CDC, NIH, and IOM have promoted the idea that the rise in autism is primarily diagnostic, while excluding or downplaying environmental and iatrogenic hypotheses. But the current data—showing accelerating prevalence, worsening severity, and growing racial disparities—make this position untenable. It is now clear that narrative closure, not causal closure, has been guiding public messaging. The refusal to explore vaccine adjuvants, prenatal toxic exposures, chronic immune activation, and regulatory policy failures reflects a broken system more committed to preserving public confidence than discovering the truth.

In a striking statement during an April 2025 Cabinet meeting, Secretary of Health and Human Services Robert F. Kennedy Jr. declared,

“By September, we will know what has caused the autism epidemic and we’ll be able to eliminate those exposures.”

The promise represents a historic shift in federal tone—marking the first time in decades that a sitting health official has committed to openly investigating all plausible causes of autism, including environmental and iatrogenic exposures.

Share