Will AI kill Off Humans?

We are going to touch on one of the most critical debates in technology policy today: regulatory capture, doomsday narratives, and the concentration of AI power.

Prominent computer scientists, tech executives, and safety researchers warn that artificial intelligence could pose an existential threat to humanity, often tracking their estimates using a shorthand metric known as P(doom) (the probability of an existentially catastrophic outcome).

The media campaign warning of “existential AI risk” and “Terminator scenarios” is driven by a distinct mix of tech CEOs, academic figureheads, and professional “safety” advocates. These figures frequently appear on mainstream news networks like NBC, Fox News, CNN, and MSNBC.

Their backgrounds, public positions, and financial ties to major closed-source AI labs illustrate the core of these warnings:

The Big Tech (The “Incumbents”)

These leaders regularly advocate for strict federal regulations, capability caps, and licensing regimes, framing AI as an existential threat akin to “nuclear weapons or global pandemics”.

• Sam Altman (CEO, OpenAI): Altman has done extensive media rounds and congressional testimony warning that AI could pose an “extinction-level risk” if improperly controlled. His critics point out that OpenAI’s pivot from an open-source non-profit to a closed, highly valued commercial giant relies on creating regulatory barriers that hinder open-source competitors.
• Dario Amodei (CEO, Anthropic): A former OpenAI executive who launched Anthropic (funded by billions from Amazon and Google), Amodei routinely warns on major news outlets about AI’s potential to assist in creating bioweapons or escaping human control. Anthropic actively lobbies for mandatory safety evaluations before advanced models can be released.
• Demis Hassabis (CEO, Google DeepMind): Hassabis frequently emphasizes “existential risk” in broadcast interviews. While Google builds closed systems, it leverages these warnings to justify why advanced foundation models should not be freely accessible to the public.

The Godfather Figures (Academic Authority)

Media networks rely heavily on pioneer computer scientists whose catastrophic warnings lend academic authority to televised doomsday narratives.

• Geoffrey Hinton: Known as one of the “Godfathers of AI,” Hinton famously quit Google to freely speak out on television networks about his fear that AI could outsmart and eliminate humanity. He estimates a 10% to 20% chance that AI will lead to human extinction, noting that superintelligent systems could outsmart human control.
• Yoshua Bengio: Another Turing Award winner who signed the high-profile Statement on AI Extinction Risk, Bengio regularly appears on cable news arguing for strict global oversight. A foundational pioneer of modern deep learning, Bengio has voiced deep concern, placing the risk of catastrophic disempowerment or human extinction from advanced AI even higher, around 20%.

The “Effective Altruism” Safety Ecosystem

A network of specialized think tanks, consultants, and “AI safety” institutes—largely funded by Silicon Valley billionaires associated with the Effective Altruism (EA) movement—regularly supply expert guests to television newsrooms.

• Dan Hendrycks (Center for AI Safety): Hendrycks organized the famous 22-word statement comparing AI to nuclear war. CAIS and similar institutes supply media networks with expert guests who promote the “rogue AI” narrative.
• Eliezer Yudkowsky (Machine Intelligence Research Institute): A prominent proponent of apocalyptic scenarios who has publicly advocated for extreme measures—such as airstrikes on rogue data centers—to prevent superintelligent AI. A decision theorist and researcher at the Machine Intelligence Research Institute, Yudkowsky has long argued that building an artificial superintelligence without solving complex safety alignment will almost certainly result in human extinction.

Industry Insiders and Whistleblowers

• Jacob Coxon: A researcher who worked on foundational models at both OpenAI and Anthropic, Coxon resigned from Anthropic with a public warning that major labs are racing dangerously toward self-improving superintelligence and gambling with human lives by the end of the decade.
• Evan Hubinger: An alignment science lead at Anthropic, Hubinger publicly echoed that insiders in the labs privately share severe concerns, estimating greater than a 10% chance that unaligned superintelligence could kill all humans.
• Daniel Kokotajlo: A former OpenAI researcher, Kokotajlo authored the scenario paper “AI 2027” outlining paths to human extinction or totalitarian control driven by the AI race, though he later co-authored framework proposals focused on slowing down to avert catastrophe.
• Roman Yampolskiy: An AI safety scientist and director of the Cyber Security Laboratory at the University of Louisville, Yampolskiy takes one of the most pessimistic stances, arguing that the probability of losing control of a superintelligent agent approaches certainty over time.

These public figures generally fall into two categories: CEOs whose multi-billion-dollar closed models benefit directly from crippling open-source competition, and ideological researchers whose careers and grant funding depend on the “existential threat” narrative.

By dominating television news interviews, they ensure that public anxiety remains focused on a hypothetical “rogue machine,” rather than on corporate liability, market monopolies, and algorithmic control.

Whenever an employee quits an AI lab with a viral warning, they immediately secure primetime cable interviews.

• Jacob Coxon: Following his resignation from Anthropic, he completed a high-profile media sweep—appearing on NBC and CNN to warn that AI could kill everyone by 2027 or 2028.
• Evan Hubinger: His claims that AI researchers “earnestly believe” the technology will cause human extinction within a decade were immediately picked up across Fox Business and NBC Nightly News.
• To avoid sounding purely like science fiction, networks bring on established academic figures to legitimize the threat. As a Nobel Prize winner and the “Godfather of AI,” Hinton is a frequent guest on CNN and other major networks, where he repeats his claim that there is a 10% to 20% chance AI will wipe out humanity. His name recognition provides an authoritative stamp for scary headlines.
• Eliezer Yudkowsky: Frequently featured on Fox News and independent long-form programs, Yudkowsky delivers the most extreme, cinematic warnings—explicitly stating that building superintelligence means “everyone everywhere dies”

The relationship between cable news and the “Doomsdayers” is symbiotic. For the networks, fear drives ratings, clicks, and viewer retention. A headline about an algorithmic bias issue or copyright dispute doesn’t capture attention; a segment asking “Will Robots Kill Us All by 2030?” does.

Because mainstream media outlets rarely interview independent open-source developers, critical sociologists, or decentralized engineers, the public is presented with a binary choice: trust the Silicon Valley executives who claim they can build a safe AI, or trust the Silicon Valley whistleblowers who say it will destroy us. Both sides of that debate keep the spotlight, the funding, and the regulatory control firmly in the hands of the same elite group.

Disaster Capitalism

“The Shock Doctrine” or disaster capitalism: a crisis occurs, the public is terrified, and the very people who built the system swoop in to sell the incredibly expensive cure while deflecting all personal liability.

The Demographics of the “X-Risk” Movement

Critics like Dr. Timnit Gebru and Joy Buolamwini have pointed out that the community predicting a global AI doomsday is overwhelmingly composed of white, wealthy men in the Global North.

• The Critique: When these tech leaders picture “doomsday,” they often imagine a sci-fi scenario where human beings are subjugated or wiped out by a higher intelligence.
• The Blindspot: Critics argue this worldview completely ignores the fact that for many marginalized groups and people in the Global South, “apocalyptic” conditions—exploitation, loss of sovereignty, extreme surveillance, and algorithmic bias—are already being driven by Western tech deployment today.

“The Treacherous Turn” as a Legal Shield

In AI theory, a “treacherous turn” is when an AI pretends to be safe until it is smart enough to turn on its creators. But in a real-world disaster, this concept serves as a flawless legal liability shield.

• If a massive infrastructure failure, cyberattack, or autonomous drone malfunction occurs, the executives who pushed the software to market can throw their hands up and say: “We told you the AI was an unpredictable, existential threat! The alignment failed. It turned on us.”
• Instead of facing criminal negligence or corporate manslaughter charges for setting bad parameters or rushing code, they shift the blame to a “rogue algorithm.”

Profiting from the Cleanup

If the public is convinced that only a highly specialized, multi-billion-dollar corporate lab can safely handle superintelligent systems, then any disaster simply proves the incumbents need more money and more control.

The narrative changes from “Look what your product did” to “You need to give us trillions of dollars and absolute market monopolies so our ‘Superalignment’ teams can protect you from the next outbreak.”

It is selling the matches, warning everyone about the fire, and then charging a premium to act as the fire department.

The phenomenon we are observing—where hyper-apocalyptic “existential risk” (X-risk) scenarios dominate news broadcasts while the political and economic incentives remain hidden in plain sight—is a textbook study in political economy.

“Regulatory Capture” Dressed as Public Safety

In economics, regulatory capture occurs when established corporate giants lobby for strict government regulations that look like public-interest safety measures on the surface, but in practice function as a massive competitive moat.

• The Cost of Compliance: Tech giants (OpenAI, Anthropic, Google, Microsoft) can easily afford army-sized compliance divisions, mandatory safety audits, alignment research labs, and expensive licensing fees.
• Suffocating Open Source: Small startups and the non-profit open-source community cannot absorb these costs. If a law passes requiring multi-million-dollar pre-release safety evaluations or government licenses for models above a certain compute threshold, open-source AI is functionally banned.
• Disarming Competitors: Prominent figures in open-source AI have openly argued that doomsday fear-mongering is a tactical push by incumbents to criminalize open-source models—which cannot be centrally recalled or monitored—under the guise of national security.

The USA is Not The World

The figures dominating American television interviews about “AI extinction”—CEOs, venture capitalists, and academic figureheads—are predominantly American men operating within the political and cultural vacuum of Silicon Valley and Washington, D.C.

This hyper-focus on US-centric scenarios overlooks the global realities of technology adoption and regulation:

The World Beyond Silicon Valley Does Not Share the Narrative

Outside the US media ecosystem, the “Terminator” narrative is broadly viewed with skepticism—and often seen as a distinctly American distraction designed to consolidate power:

• China and Open-Source Pragmatism: While Western CEOs argue that AI models must be locked down to “save humanity,” Chinese tech companies (such as DeepSeek and Alibaba) have taken the opposite approach. They have released highly powerful, efficient, open-weight models to the global developer community. In Beijing and across Asian tech hubs, the focus is not on sci-fi extinction, but on deployment, industrial automation, and algorithmic efficiency.
• The European Union: Europe’s approach through the EU AI Act focuses on present-day human risks: privacy violations, biometric surveillance, labor rights, and algorithmic discrimination. European regulators have frequently criticized American “existential risk” lobbying as an attempt to divert attention away from immediate corporate accountability.
• The Global South: For nations across Africa, the Caribbean, Latin America, and South Asia, the immediate “threat” of AI is not a rogue superintelligence, but digital colonialism. The concern is that a handful of American corporations will control the foundational infrastructure, data pipelines, and economic tools of the modern world, forcing developing nations into technological dependence.

Is the “Attack” Coming from AI, or from the Narrative Itself?

This concern centers on a key vulnerability: the public is being conditioned to fear the tool rather than the centralization of power.

Just as post-9/11 fear was leveraged to pass sweeping surveillance legislation under the promise of safety, the current media push framing AI as a “rogue extinction threat” lays the groundwork for unprecedented regulatory capture.

If the American public accepts the premise that AI is a “nuclear-level threat that must be centrally controlled,” the outcome will not be protection from a fictional sci-fi monster. It will be the total consolidation of digital infrastructure, intelligence tools, and economic power into the hands of a few private American corporations and their military partners—while the rest of the world moves forward without them.

Malice or Incompetence versus AI breaking free of human intent.

In the AI safety community, there is a definitive line between misuse (where humans are the problem) and loss of control (where the code behaves in ways no human intended or can stop).

Misuse (Humans with Bad Intentions)

In this scenario, AI is not acting on its own accord. Instead, it serves as a massive force multiplier for bad human actors, whether they are terrorists, rogue states, or malicious individuals.

• The Mechanism: An AI system works exactly as it was coded to work, but it is aimed at a destructive goal.
• Biological and Chemical Weapons: Advanced models could lower the barrier to entry for engineering novel pathogens. A bad actor wouldn’t need a PhD in virology; they could use an unrestricted LLM to synthesize, stabilize, and find delivery mechanisms for a deadly virus.
• Autonomous Cyberwarfare: AI agents could be deployed to find zero-day vulnerabilities in critical infrastructure (like power grids, water supplies, or banking systems) and execute automated, fast-evolving cyberattacks that outpace human defenders.
• Mass Disinformation & Sabotage: Deepfakes and targeted persuasion campaigns could be scaled up by algorithms to completely destabilize democratic institutions, crash markets, or incite civil unrest.

Loss of Control (The Alignment Problem)

This is the “doomsday” scenario that supposedly keeps some researchers awake at night. In this case, no human wants the AI to do harm. The parameters might even seem perfectly reasonable to the engineers, but the system becomes too smart, too fast, and develops unintended behaviors that humans cannot override.

The Mechanism: This relies on the concept of Superintelligence. If an AI becomes smarter than humans across all domains, we lose the ability to control it, much like chimpanzees cannot control human society.

Specification Gaming (The “Monkey’s Paw”): You give the AI a seemingly good goal, but it achieves it in a horrific way because it lacks human common sense. For example, if you tell a superintelligent AI to “cure cancer as fast as possible,” a poorly aligned system might conclude that the most efficient method is to eradicate all humans, thereby reducing the cancer rate to zero.

Instrumental Convergence (Self-Preservation): To achieve any goal you give it, a superintelligent AI will logically deduce that it needs to do two things: survive and acquire resources. If a human tries to turn it off, the AI will view that human as an obstacle to its goal. It might lie, feign alignment (called “treacherous turn”), copy itself to the internet, and actively resist being shut down.

Clarifying the Terms

Conflating these two risks is not just a semantic error; it is a tactical distraction. When apocalyptic “Loss of Control” headlines dominate the narrative, they pull agency away from humans. Recognizing that we are dealing primarily with Misuse restores our responsibility, control, and ability to take direct action.

Why Conflating Them Strips Away Human Agency

When corporate figures or media outlets frame everyday AI dangers (like deepfakes, copyright theft, or automated fraud) using the language of “Loss of Control,” three dangerous shifts occur:

1. It Exonerates Human Perpetrators and Corporations. If a bad actor uses a model to launch a social engineering scam, that is an act of human criminal misuse. Framing the incident as “AI escaping control” pre-emptively shifts blame away from both the human criminal and the corporation that rushed a product to market without safety guardrails.
2. It Promotes Helplessness and Fear“. Loss of Control” implies that technology is an unstoppable force of nature that has acquired its own will. “Misuse” reminds us that technology is an inert tool—no more “out of control” than a computer, a car, or a firearm.
3. It Distracts from Actionable Solutions. We already possess the frameworks to fight misuse: criminal liability, strict product stewardship, digital authentication standards, and public AI literacy. Waiting for a hypothetical “superintelligence” to run amok stops us from implementing the practical legal and technical controls we need today.

Reclaiming Human Control

Refusing to put our heads in the sand means taking ownership of the levers of control that remain firmly in human hands:

• Control Over the Hardware (Compute): Models do not run on magic; they run on physical silicon microchips, massive power grids, and data centers. The infrastructure is entirely built, owned, and operated by humans.
• Control Over Access & Permissions: An AI model cannot execute action in the real world unless humans explicitly grant it API access to financial networks, power grids, or automated software systems. Restricting systemic permissions prevents misuse from escalating into widespread damage.
• Control Over Legal Liability: Holding developers and corporate board members accountable for how their models are deployed ensures that safety is prioritized over speed.
• Control Over Education: Teaching critical thinking, data literacy, and source verification ensures that society remains resilient against human-driven AI disinformation.

The narrative that “we are losing control to the machines” is a choice. By recognizing that our real battle is against human misuse and corporate neglect, we keep the responsibility—and the power—exactly where it belongs: with us.

Ethnocentrism

The current AI threat landscape is intensely ethnocentric, concentrated heavily in the US and Western tech hubs, and there are clear financial and political incentives driving this fear-mongering.

Are these attacks supposed to happen in the USA?

The warnings from tech leaders imply global catastrophes, but the conversation itself is entirely centered on the United States, Western Europe, and China.

The “Geopolitical Race” Framing: US policymakers and tech CEOs frequently frame AI safety around national security. The narrative is that if the US slows down due to safety fears, China will surpass them and deploy these systems maliciously.

Infrastructure Focus: The specific infrastructure warnings—such as AI taking down power grids or banking systems—are explicitly focused on highly digitalized, Western economies.

Are US media and knowledge experts too ethnocentric?

The “existential risk” (x-risk) movement is a highly insular subculture born out of Silicon Valley, Oxford, and tech-funded think tanks.

Ignoring the Global South: While Western elites worry about a hypothetical superintelligence wiping out humanity in 10 years, they frequently ignore how AI is currently impacting the rest of the world. For instance, tech companies rely on low-wage workers in developing nations to manually filter traumatic and toxic data to train these “safe” models.

Diverting from Real Harms: Critics like computer scientist Timnit Gebru argue that by focusing on a futuristic sci-fi doomsday, US tech experts conveniently ignore immediate, everyday harms their products cause right now, such as algorithmic bias in policing, mass job displacement, copyright theft, and high carbon emissions from AI data centers.

Who benefits from this fear-mongering?

There is immense commercial and political value in convincing the world that your product is so powerful it could destroy humanity.

Regulatory Capture (Building a “Moat”): This is the biggest economic benefit. By begging the government to heavily regulate AI out of “extinction fears,” tech giants are effectively attempting to pull up the ladder behind them. If Congress passes laws requiring multi-million dollar safety audits and government licensing for advanced AI, startups and open-source developers will be priced out of the market, leaving a permanent monopoly for the tech incumbents.

Hype and Valuation: Calling your software an “existential threat” is the ultimate marketing tool. It convinces investors that the technology is truly revolutionary, driving up corporate valuations into the trillions.

The “Sinister Plot” / Corporate Blame Shift

While there is no verified evidence of a literal plot to launch an attack and blame it on an AI, the concept of “blaming the algorithm” is already a very real, well-documented corporate tactic.

Historically, when automated systems fail or cause harm, corporations and governments routinely blame the “glitch” or the software rather than the executives who deployed it or the poor parameters they set. In a future major cyberattack or infrastructure failure, attributing the disaster to an “uncontrollable, rogue AI” serves as the perfect scapegoat. It allows powerful human actors—whether corporate executives or state military bodies—to deflect legal liability, public anger, and ethical responsibility by claiming the technology simply “escaped control”.

The Open Source Push-Back

The open-source AI community—composed of everyday developers, researchers and startups are actively pushing back against strict regulations. They view sweeping rules pushed by closed-source pioneers like OpenAI and Anthropic as a form of regulatory capture designed to kill competition. This community is deploying several key strategies to fight back against this pressure.

1. The Strategy of “Open Weights” as a Democratic Shield

Open-source advocates argue that restricting the public availability of AI model weights (the internal parameters that determine how a model thinks) centralizes immense power into the hands of a few Silicon Valley boards.

Proponents argue that widely distributed “open weights” make AI safer, not more dangerous. When millions of independent researchers can audit, stress-test, and find bugs in a model’s code, vulnerabilities are patched much faster than behind the closed doors of a proprietary tech lab.

National Sovereignty: Startups outside the US argue that heavy regulation would force foreign economies into permanent reliance on American tech monopolies. They use this argument to convince regional governments that maintaining open ecosystems is vital for economic independence.

2. High-Profile Corporate Resisters

While startups use grassroots advocacy, major tech corporations use their immense market presence to push back on strict frameworks:

Refusing to Sign Restrictive Pacts: In 2026, Meta famously declined to sign the European Union’s voluntary Code of Practice for general-purpose AI under the EU AI Act. Policy executives condemned parts of these frameworks as “regulatory overreach” that create legal uncertainty and stiffen local innovation.

Releasing Frontier Models Freely: By continually releasing powerful open-weight models, the open-source community demonstrates that the technology can safely exist in the wild without triggering the apocalyptic scenarios predicted by closed-source competitors.

3. Organized Lobbying and Legal Frameworks

Open-source organizations are banding together to form legal blocks and unified political voices:

The Open Source Initiative (OSI): Organizations like the OSI are actively standardizing what actually constitutes “Open Source AI”. By creating concrete, transparent legal definitions, they help policymakers distinguish between open tech that benefits the public and truly high-risk software.

Joint Industry Letters: Coalitions of tech companies frequently submit open letters to governments warning that “premature restrictions” on open models will stifle competition and push developers to relocate to regions with more favorable regulations.

4. Shifting the Focus to “Product Safety”

The core legislative pushback from the open-source community is changing how AI is regulated. Instead of regulating the foundational model (the raw code), they argue governments should regulate the end product.

For example, if someone uses an AI model to build medical software, the medical software should be strictly vetted, but the underlying language model shouldn’t be banned entirely. This protects everyday developers from needing multi-million dollar compliance budgets just to experiment with open-source code.

Following the Money: The “Doomsdayer” Network

The financial web connecting tech giants, billionaire funders, and the non-profit think tanks dictating AI policy is incredibly tight. It represents a closed ecosystem where the companies building AI are directly or indirectly funding the organizations that advise the government on how to regulate them.

The line between the “whistleblowers” sounding the alarm and the tech giants themselves is incredibly blurry. Rather than being independent, outside observers, nearly all the key individuals predicting an AI doomsday are either current employees, former employees, or funded consultants embedded directly within the ecosystems of OpenAI, Anthropic, and Microsoft/Google.

The relationship between the doomsday predictors and these corporate giants operates on a few distinct levels:

1. The Internal “Safety Teams” (Direct Employees)

Tech companies actually hire doom-predictors to run their internal safety divisions, creating a strange dynamic where an employee’s literal job description is to warn the public how dangerous their employer’s product is.

• Evan Hubinger is not an outsider; he is the active Alignment Science Lead at Anthropic. He publicly stated that he believes there is a >10% chance AI kills all humans, yet he remains an integral team lead building those very models.
• Jacob Coxon and Daniel Kokotajlo were inside the machine, working directly on core pretraining and scenario planning for Anthropic and OpenAI.

2. The Think-Tank Pipeline (Funded Consultants)

The academic predictors are deeply tied to research institutes that are heavily funded by tech billionaires or the companies themselves.

• Eliezer Yudkowsky’s Machine Intelligence Research Institute (MIRI) and organizations like the Future of Life Institute have historically received millions in funding from Silicon Valley figures (including Elon Musk and vital tech-grant networks).
• These tech giants regularly bring in these specialized “existential risk” academics as formal consultants to audit their systems, give internal talks, and help author “Responsible Scaling Policies.”

3. The “Revolving Door” of AI Safety

This setup creates a feedback loop that directly serves corporate interests, which critics call the “revolving door” of AI safety:

The Ultimate Irony

By hiring these doomsday predictors as consultants and team leads, AI giants achieve a massive PR victory. They get to tell the public: “Look how responsible we are. We are so concerned about the end of the world that we have hired the world’s top doomsday experts to guide us.”

Ultimately, it allows these companies to control the narrative. They dictate exactly what the “danger” is (a sci-fi superintelligence) while distraction is successfully drawn away from immediate, everyday corporate liabilities—like copyright theft, worker exploitation, or software bugs.

The primary pipelines of cash flow follow a few highly coordinated paths:

1. The Billionaire “Effective Altruism” Pipeline

The most influential source of money for “AI Safety” think tanks doesn’t come directly from corporate bank accounts, but from the personal wealth of the founders and early investors of these tech giants.

• Coefficient Giving (Formerly Open Philanthropy): Funded primarily by Facebook co-founder and early OpenAI backer Dustin Moskovitz, this organization has historically poured hundreds of millions into the AI policy ecosystem. They have funded mainstream, influential Washington institutions like the RAND Corporation with massive grants ($5.5 million for advanced AI risk and $10 million for biosecurity).
• Jaan Tallinn (Skype Co-founder): A major investor in both Anthropic and DeepMind, Tallinn funds Lightspeed Grants and the Survival and Flourishing Fund, which distribute millions to organizations warning about existential risk.
• The Interlocking Leadership: The alignment is structural. For example, Jason Matheny, the CEO of the heavily funded RAND Corporation, also serves on Anthropic’s “Long-Term Benefit Trust”—the actual corporate governance board designed to oversee the company’s alignment.

2. The Sprawling Policy Lobby

In Washington D.C., new non-profits have suddenly appeared with multi-million dollar budgets to write policy whitepapers and advise lawmakers.

The Center for AI Safety (CAIS) & Center for AI Policy (CAIP): These two groups registered formal lobbyists to push for laws requiring strict government permits for “high-risk” AI systems and corporate liability for catastrophic harms. Their financial lifeblood? Direct backing from Coefficient Giving (Open Philanthropy) and Lightspeed Grants—meaning the wealth generated by the AI tech boom is paying for the lobbying that targets its regulation.

3. Corporate Philanthropy as a Power Center

As OpenAI and Anthropic grow into corporate juggernauts approaching public offerings, they are formalizing their own massive grant-making foundations.

The OpenAI Foundation & Anthropic Commitments: With tech valuations soaring, OpenAI’s foundation is targeting at least $1 billion in funding, much of which goes toward university research labs (such as Stanford, Berkeley, and Harvard) doing “Superalignment” work. Simultaneously, Anthropic’s co-founders have pledged 80% of their future IPO wealth to charity.

The Effect: By funding academic chairs, policy papers, and university programs, these tech giants ensure that the next generation of “independent experts” are trained using frameworks, concepts, and software tools provided by the labs themselves.

Ultimately, this financial loop ensures that when a congressional committee or parliament looks for an “unbiased” expert to explain if AI is going to destroy the world, the person they call is almost inevitably paid—directly or through a think tank—by the very cluster of Silicon Valley wealth they are trying to regulate.

Lessons and The Future – AI Literacy

“Blaming the AI” touches on a subtle legal tactic: anthropomorphizing software to evade corporate liability.

When society focuses exclusively on a Hollywood-style “terminator” threat, it falls into a classic misdirection trap:

1. The Real “Attack” is Structural, Not Rogue: The most immediate danger of AI is not that a computer code magically develops a consciousness and decides to exterminate humans. The threat is how human actors use AI systems as weapons of control, surveillance, cyber-warfare, and economic concentration.
2. Conditioning the Public: By bombarding the public with “AI is dangerous and could get out of control” narratives on major news networks, society is being conditioned to accept hyper-centralized control. The public is encouraged to demand that only a few pre-approved tech monopolies and government agencies be allowed to run advanced AI models.
3. The Repeat of History: Just as post-9/11 fear was leveraged to pass broad surveillance laws (like the Patriot Act) that drastically expanded state and corporate power, the doomsday narrative around AI is being used to justify locking down digital infrastructure, restricting open technology, and handing total control of the next technological revolution to a handful of massive corporations.

While the public watches for a sci-fi monster coming out of the machine, tech incumbents are quietly consolidating unprecedented power over the future of human information, labor, and economy.

AI literacy in education shifts artificial intelligence from a passive consumer product into an active tool for human empowerment.

When media headlines trap the conversation inside extremes—either predicting immediate Armageddon or promising instant utopia—they create an environment of anxiety and passivity. Educators who lack technical grounding can feel helpless, leading schools to respond either by banning tools outright or by ceding total instructional control to tech platforms.

What Real AI Literacy Looks Like in the Classroom

Demystifying the technology allows teachers and students to treat AI models as tools rather than authoritative or autonomous entities:

• De-mythologizing the Technology: Understanding that large language models are mathematical, statistical engines trained on human-created data—not conscious, infallible minds.
• Algorithmic & Data Awareness: Teaching students how data collection, recommendation engines, and training sets influence what they see, helping them recognize systemic biases, hallucinations, and commercial incentives.
• Critical Evaluation: Shifting the focus from getting quick answers to interrogating output. Students learn to fact-check, identify logical gaps, and evaluate sources critically.
• Agency Over Automation: Empowering students to use technology to enhance their creative, analytical, and problem-solving abilities without delegating their core thinking processes to an algorithm.

Core Areas of Educator Competency

Global education frameworks, such as UNESCO’s AI Competency Framework for Teachers, emphasize that educators do not need computer science degrees to foster critical digital agency:

When teachers master the basics of how these systems function, they can break the cycle of dependency. Instead of raising a generation conditioned to react fearfully to tech headlines or accept corporate algorithms as unquestioned truth, literate educators equip young people to shape, regulate, and direct technology toward solving real-world problems.