OpenAI has an unsustainable business model ('A+ technology with a C- business model') and is at high risk of financial failure or a discounted sale due to its massive burn rate and poor monetization.
Effective AI safety cannot be left to competing private labs and requires robust government regulation, including a new federal agency and international treaties involving China.
The motivations of AI leaders—Hassabis's science, Altman's power, Musk's legacy—are critical drivers of the industry's trajectory, talent retention, and competitive dynamics.
U.S. chip export controls against China have been ineffective due to loopholes like third-country cloud computing access, allowing China to remain a close competitor just six months behind the U.S. frontier.
The U.S. government is increasingly intervening in the tech sector, taking equity stakes in strategic companies like Intel and imposing heavy-handed controls on frontier AI labs.
2014
DeepMind is acquired by Google, a move Malaby characterizes as a strategic way to secure massive R&D funding for its London-based operations. A condition of the deal was an ethics oversight board.
Summer 2015
DeepMind holds an AI safety summit at SpaceX, hosted by Elon Musk, in an attempt to create a unified safety oversight board.
Late 2015
Elon Musk co-founds OpenAI, a direct competitor to DeepMind, signaling the start of a competitive race rather than a unified safety effort.
Late 2022
OpenAI launches ChatGPT, which Demis Hassabis of DeepMind reportedly views as a 'declaration of war.'
January (Recent Year)
Malaby publishes articles predicting a 50% chance of OpenAI's financial collapse within 18 months, citing an internal projected burn rate of $660 billion over five years.
Post-2025 (Recent/Hypothetical)
The emergence of powerful models like Anthropic's 'Mythos' prompts a significant shift in the U.S. government's stance from laissez-faire to active regulation.
▶OpenAI's Unsustainable Bubble
Malaby argues that OpenAI's finances are extremely precarious due to a massive burn rate ($660 billion projected over five years), a high valuation ($852B) unsupported by revenue, and an inability to monetize its 900 million users. He believes the company is caught in a competitive squeeze and faces a 50% chance of failure or a forced sale by mid-2025.
Investors should be highly skeptical of OpenAI's private valuation and view its potential IPO not as a sign of strength, but as a desperate move born from an unsustainable business model, similar to WeWork.
▶The AI Arms Race and Geopolitics
Malaby portrays a fierce competition not just between U.S. labs like DeepMind, Anthropic, and OpenAI, but also on a global scale with China, which remains only six months behind the U.S. despite ineffective export controls. He highlights how this race dynamic complicates safety efforts and necessitates international agreements.
The ineffectiveness of current U.S. chip controls suggests that supply-side restrictions are insufficient, and analysts should focus on the geopolitical necessity of diplomatic engagement with China on AI safety to mitigate global catastrophic risks.
▶The Inevitability of Government Intervention
Malaby asserts that the immense power of frontier AI models, exemplified by the hypothetical 'Mythos' model, has forced a reversal of the U.S. government's laissez-faire approach. He advocates for a new regulatory body akin to the FDA and notes the government is already taking unprecedented steps, such as taking equity in private firms and mandating customer-by-customer approval for model access.
The trend towards heavy-handed regulation is accelerating, creating significant compliance and operational risks for frontier AI labs, while also potentially creating a 'too important to fail' moat for companies that align with government strategic interests.
▶The Personalities and Motivations of AI LeadersJul 2026
Malaby's analysis heavily focuses on the personalities and motivations of key AI leaders, contrasting Demis Hassabis's deep scientific drive with Sam Altman's pursuit of power and Elon Musk's quest for legacy. He argues these personal motivations are critical in shaping company strategy, talent recruitment, and the overall trajectory of the AI race.
Talent retention and long-term strategic focus in AI may be more closely tied to the perceived scientific integrity of leadership (like Hassabis and Amodei) than to market hype or financial incentives, making leadership analysis a key factor in assessing a lab's viability.