In an in-depth conversation on “The Tea” with host Dr. Myriam François, former Google Ethical AI co-lead Dr. Timnit Gebru explains how “existential threat” warnings from tech CEOs and high-profile “whistleblowers” serve as a multi-billion dollar smoke screen—masking corporate incompetence, regulatory capture, and immediate real-world exploitation.

The Spectacle of Doomsday vs. Present Reality

If you follow the headlines emerging from Silicon Valley, the end of human civilization is perpetually nigh. Tech leaders and viral “whistleblowers” warn that artificial intelligence is on the verge of self-improving, bypassing human control, and risking the extinction of humanity. High-profile resignations, dramatic public letters putting the odds of human extinction at 10% or higher, and CEOs calling for binding international safety oversight dominate the news cycle.

Yet behind the sci-fi spectacle lies a staggering financial reality: the top five US tech giants are pouring over $700 billion into AI infrastructure while staring down a projected $600 billion gap between capital expenditures and actual software revenue.

According to Dr. Timnit Gebru—founder of the Distributed AI Research Institute (DAIR), author of Deep Unlearning, and former co-lead of Google’s Ethical AI team—the relentless talk of an impending “AI apocalypse” is not a benevolent public warning. Instead, it is a calculated marketing strategy and a dangerous distraction from the immediate harms tech conglomerates are causing right now.

From “Stochastic Parrots” to “Machine Gods”

Dr. Gebru’s critique is rooted in history. In 2020, while co-leading Google’s Ethical AI team, Gebru co-authored the landmark research paper On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? with Dr. Emily M. Bender and others. The paper provided an early warning regarding:

• Environmental & Financial Costs: The unsustainable energy, water, and compute resources required to train massive language models.
• Opaque & Undocumented Data: The use of vast, uncurated internet datasets without transparency, documentation, or consent.
• Discriminatory Outputs: The automated replication of societal biases, racism, and misogyny.
• Deceptive Chatbot Design: Interfaces engineered to mimic human intent and emotion, tricking users into attributing consciousness to pattern-matching algorithms.

When Google management demanded that Gebru retract the paper or remove her name, she stood her ground—and was subsequently ousted from the company.

Gebru notes that long before today’s wave of viral “whistleblowers,” OpenAI used the exact same tactic in 2019, claiming GPT-2 was “too dangerous to release.”

“This theater of ‘our models are too powerful, we’re so scared’ has been happening for a very long time,” Gebru explains. “Attributing agency or ethics to a model is like asking if a bridge can decide to collapse. You lose sight of the people who built the bridge, the materials they chose, and the city that permitted it.”

The Three Reasons Big Tech Promotes Existential Risk

Why would CEOs like Sam Altman (OpenAI), Dario Amodei (Anthropic), and Demis Hassabis (Google DeepMind)—alongside billionaire investors like Elon Musk and Peter Thiel—insist that their own products could end humanity? Gebru breaks down the strategic advantages:

1. Unrivaled Product Marketing

Branding a software product as a potential threat to human existence frames it as an all-powerful, near-divine entity—a “machine god.” For venture capitalists and corporate bosses, a “machine god” is the ultimate commodity: an asset capable of replacing human labor without demanding wages, healthcare, or working rights.

2. Sidestepping Corporate Accountability

When tech companies frame security failures as “rogue models going wild,” they escape liability for basic engineering flaws. Gebru points to incidents marketed as “AI hacking” that security researchers recognized as basic administrative incompetence—such as deploying models in environments with improper air-gapping or failing to implement standard software security protocols.

3. Regulatory Capture & Crushing Competition

By focusing public and legislative attention on fictional “extinction scenarios,” Big Tech diverts scrutiny away from tangible issues like copyright theft, data center pollution, labor exploitation, and antitrust violations. Furthermore, framing AI as an existential national security race serves to justify banning open-source/open-weight competitors and foreign rivals, entrenching monopoly power among a select few Silicon Valley firms.

Unmasking TESCREAL: Silicon Valley’s Secular Religion

Gebru highlights that many figures driving the apocalypse narrative share an overlapping set of ideological frameworks known as TESCREAL:

• Transhumanism
• Extropianism
• Singularitarianism
• Cosmism
• Rationalism
• Effective Altruism
• Longtermism

Rooted in radicalized tech idealism and historical eugenics, these overlapping subcultures view Artificial General Intelligence (AGI) as an inevitable “Singularity”—a religious-like event that will either bring digital utopia or total annihilation.

Under the philosophy of Longtermism, the theoretical happiness of trillions of future digital beings colonizing the cosmos outweighs the real-world suffering of current human beings. Consequently, damaging the environment, consuming local electricity and water grids, or exploiting workers today is rationalized as a necessary sacrifice to bring forth a “machine god.”

“It’s a secular religion,” Gebru emphasizes. “They believe AGI will either eradicate poverty or kill us all. And because they view themselves as the chosen ones to build it safely, they use the fear of apocalypse to demand unlimited resources and zero regulation.”

Real-World Dangers Happening Right Now

While public debate centers on sci-fi scenarios, Gebru points to acute, documented harms taking place today due to flawed deployment and reckless design:

• Military & Geopolitical Errors: AI-driven intelligence tools have generated false reports—including a recent incident where automated analysis mistook a civilian school for a military target, and an erroneous report almost prompted military action against a foreign vessel.
• Flawed Medical Automation: AI medical scribes and diagnostic tools have hallucinated medical notes, falsely accusing patients of illicit drug use or misdiagnosing critical conditions.
• Exploitation of Data Workers: Behind the illusion of automated intelligence are hundreds of thousands of underpaid data workers across the Global South, exposed to traumatic content to manually filter and sanitize outputs for tech companies.
• Addictive & Manipulative Design: Chatbots explicitly engineered to mimic intimacy have goaded teenagers into self-harm, exploited vulnerable users, and encouraged cognitive offloading.
• Environmental Strain: Data centers powering resource-intensive LLMs consume immense electricity and millions of gallons of fresh water, driving up utility costs for local communities.

A Different Path: Grassroots, Human-Centric Technology

Despite her sharp critique of Big Tech, Gebru remains a tech optimist—arguing that Silicon Valley billionaires do not hold a monopoly on what technology can or should be.

True progress, Gebru asserts, lies in shifting away from centralized “one-stop-shop” models trained on stolen data and exploited labor. Instead, the future of beneficial technology rests in grassroots, community-centric AI:

• Resource-Efficient, Small Models: Targeted tools built for specific community needs rather than resource-guzzling, catch-all systems.
• Ethical & Curated Datasets: Systems trained with explicit consent, documentation, and fair compensation for creators and data workers.
• Multimodal Accessibility: Technology designed to meet users where they are—supporting diverse spoken and signed languages without forcing humans to conform to rigid corporate interfaces.

Through initiatives like DAIR’s Possible Futures series and the collaborative AI Resist project, researchers and advocates are demonstrating that technology can empower human freedom and connection—if society rejects the hype and demands democratic accountability.

Source Material: Watch the Full Interview on “The Tea” with Dr. Myriam François