The INSANE Truth About OpenAI

Overview

The video argues that OpenAI’s rise was driven by a mission to build AI “safely” and keep it open to the public. However, it claims internal culture, funding pressures, and power struggles quickly shifted the company toward secrecy, commercialization, and conflict between product speed and safety.

It presents OpenAI’s trajectory as a sequence of changing priorities—from an “open nonprofit safety mission” to a scaled, closed, profit-driven product—punctuated by internal conflict and extreme power reversals.

How OpenAI Was Formed (Origin Narrative)

The video begins with a “how OpenAI was formed” narrative, describing:

• Sam Altman’s Silicon Valley rise, including connections to Paul Graham and Y Combinator
• Elon Musk’s concerns about AI safety and monopoly risk from Google
• A 2015 dinner involving major tech figures

The group, as framed by the video, imagines building an AI company to rival Google. OpenAI is portrayed as founded as a nonprofit to avoid profit motives, and the video emphasizes “openness” as central to the brand—sharing work openly rather than monetizing it.

Early Goals: AGI Hope and Existential Dread

In its early portrayal, OpenAI is described as:

• Small and experimental
• Aiming for AGI (artificial general intelligence)

The video highlights a dual mindset:

• Utopian hopes (e.g., solving disease, poverty, jobs)
• Existential dread about unintended consequences if superintelligent systems pursue misaligned goals

It also claims OpenAI’s worldview blended sincere belief with investor and public-reputation dynamics.

Transformers and the Acceleration Toward GPT

A key inflection point is the transformer architecture (“Attention Is All You Need”), originally developed by Google engineers. The video claims:

• Google approached its chatbot cautiously and avoided public release, citing reputation and business risk, especially regarding search advertising.
• OpenAI adopted transformers quickly, enabling rapid progress toward GPT models.

The video also argues OpenAI’s momentum depended on large-scale training data from the internet, raising a tension between:

• Copyright concerns
• The “ask forgiveness, not permission” approach

It further claims OpenAI withheld some releases (e.g., GPT-2) due to safety and malicious-use concerns—framing this as an early sign that “open” would not hold.

Funding, Control, and the Shift Toward Commercialization

Another major turning point is described as funding and control:

• Elon Musk resigns from OpenAI’s board, allegedly after pushing for influence (including merging OpenAI with Tesla) and losing to the board.
• The video claims Musk partially withdraws committed funds, contributing to a financial crunch.
• It describes OpenAI’s controversial switch from nonprofit to for-profit, along with commercial licensing of its technology.
• This is framed as a betrayal of the original democratizing/open mission.

The video also presents Microsoft’s partnership/investment as crucial to scaling compute, while arguing OpenAI became more entangled with large tech rather than serving as an alternative.

ChatGPT as a Global Catalyst (Nov 30, 2022)

The release of ChatGPT (Nov 30, 2022) is presented as a global accelerant:

• The video claims ChatGPT went viral faster than employees expected, becoming the fastest app to reach 100 million users.
• It asserts that internal safety concerns existed—reportedly with some safety staff not fully aware.
• It highlights risks such as:
  • Misuse (cybercrime, vulnerability discovery, weaponization)
  • Hallucinations

After launch, the video claims:

• Microsoft invests an additional $10B
• OpenAI accelerates commercialization (paid tiers and enterprise sales)

It then describes a widening internal divide between “product” and “safety”, including:

• Employee accusations that OpenAI prioritized profit over safety
• Restrictive agreements
• Resignations of key safety personnel
• Talent departures to build a competitor, Anthropic, positioned around safety

The 2023 Leadership Coup and Rapid Reversal

The most dramatic segment covers OpenAI’s 2023 leadership coup. The video claims that:

• Less than a day after Sam Altman’s public remarks about the technology’s impact, he was fired via board call on Nov 17, 2023.
• The reasons are described as vague, including alleged issues with candid communication and trust.

It offers competing theories:

• Conflicts with board members
• Altman being distracted by side projects (e.g., Worldcoin, AI chip fundraising, universal basic income framing)
• The most concerning possibility: that internal safety concerns may have motivated Altman’s board opponents

The video portrays a rapid “employee revolt”:

1. Greg Brockman resigns in solidarity.
2. Employees sign petitions threatening to leave unless Altman and Brockman return.
3. Microsoft offers to hire OpenAI employees.
4. Ilia Sutskever (described as one of the pushing board members) reportedly reverses course by joining a petition, forcing renegotiations.
5. By Tuesday, Altman and Brockman are reinstated; board members involved are removed.
6. The video ultimately frames Altman as becoming more powerful after the collapse.

Broader Implications and Current Context

The video widens to broader concerns, including:

• Risks of AI voice/personation, including references to:
  • “Her”-style voice mode
  • Scarlett Johansson’s lawsuit
• AI accelerating creative and workplace disruption earlier than expected
• Erosion of trust online as AI-generated media becomes indistinguishable

It ends with market context:

• OpenAI’s large 2025 funding round (reported as $40B at a $30B valuation)
• Intensifying competition (notably DeepSeek)
• Claims that OpenAI accused DeepSeek of IP theft, while acknowledging the irony that OpenAI itself relied on scraping data

Core Claim / Thesis

Overall, the video’s central report/opinion is that OpenAI’s story is a chain of shifting priorities—moving from an “open nonprofit safety mission” to a scaled, closed, profit-driven product—under competitive and governance pressures, accompanied by internal conflict and dramatic reversals of power.

Presenters / Contributors

• Sam Altman (subject; not the presenter)
• Elon Musk (subject; not the presenter)
• Greg Brockman (subject; not the presenter)
• Ilya Sutskever (spelled “Ilia” in subtitles; subject; not the presenter)
• Paul Graham (subject; not the presenter)
• Larry Page (subject; not the presenter)
• Helen Toner (subject; not the presenter)
• Scarlett Johansson (subject; not the presenter)
• Vladmir Putin (quoted; subject; not the presenter)
• The video narrator / host (creator not named in the subtitles)
• Michael / “AI” voice mode (character reference only; no additional named contributors beyond what’s listed above)