Ilya Sutskever

“Author of "We're Moving From The Age Of Scaling To The Age Of Research"”

All major AI labs are simultaneously converging on world models research—a shift from pattern matching and token prediction to causal simulation and state prediction—marking a significant paradigm change in AI development.

“Described emotions as value functions in a podcast discussion about moving from pattern matching to planning and simulation.”

“A cofounder of OpenAI and its former chief scientist who left the company this year to boost his focus on safe AI and founded his own company.”

“Mentioned as example of researcher who raised billions in venture capital for a startup.”

“Ilya Sutskever, superalignment co-lead at OpenAI, left the company earlier this year amid organizational restructuring.”

“OpenAI co-founder and chief scientist who compiled seventy pages of evidence about Altman's deceptive practices and led the board's attempt to fire him in 2023.”

“Ilya Sutskever founded Safe Superintelligence Inc. (SSI), a new AI lab that has raised approximately $8 billion to develop frontier AI with a focus on safety.”

“Founded Safe Superintelligence (SSI), reportedly planning to release its first model in August 2026.”

“Ilya Sutskever, cofounder and CEO of Safe Superintelligence Inc., announced a partnership with NVIDIA to scale up SSI's research using NVIDIA's Vera Rubin platform.”

“Former OpenAI co-founder who led superalignment team that Aschenbrenner joined, later left to start his own company.”

“Quoted on SSI's research and need for scaling compute via Nvidia Vera Rubin platform investment.”

“Signed the Pacing the Frontier statement emphasizing need for unprecedented measures for future AI power.”

“AI researcher who offered a periodization of modern AI history from 2012-2020 as research phase, 2020-2025 as scaling, and return to research with big computers.”

“Former OpenAI co-founder and alignment lead who founded Safe Superintelligence and announced partnership with Nvidia.”

“Founded SSI, which emerged from two years of stealth development with a $5 billion investment from NVIDIA to build safe superintelligence.”

“Former OpenAI chief scientist who founded the secretive AI startup Safe Superintelligence, which raised up to $5B in investment from Nvidia.”

“OpenAI cofounder whose stake in the company is valued at around $7 billion, as disclosed in testimony during OpenAI's trial.”

“Former OpenAI cofounder now predicts AGI could arrive between 2030 and 2045, much wider range than earlier estimates.”

“OpenAI co-founder described by Musk as the #1 person in computer vision, above Karpathy.”

“Testified as high-profile witness in Musk's lawsuit against OpenAI.”

“Testified in Musk v. Altman trial that Sam Altman is not to be trusted.”

“OpenAI executive whose personal stake in the company is worth approximately $7 billion.”

“Testified in Elon Musk lawsuit against OpenAI, disclosing equity stake valued at nearly $7 billion.”

“In a 52-page memo to OpenAI's board, stated that Sam Altman exhibits a consistent pattern of lying, undermining executives, and pitting them against one another.”

“Former OpenAI chief scientist now founder of his own AI startup; listed as angel investor in Cerebras.”

“Former OpenAI researcher who believed deep learning would gradually change lives and coined the internal meme 'Feel the AGI'.”

“Co-led OpenAI's Superalignment team with Jan Leike; wrote about scaling alignment techniques to superintelligence.”

Continual learning—the ability for models to learn new information over time without catastrophic forgetting—is essential for building more intelligent, efficient systems and addresses a critical limitation of current neural networks.

“Noted that humans rely on continual learning naturally and are very flexible to changing data.”

Leading AI researchers including Ilya Sutskever and Andrej Karpathy are revising AGI timelines and expressing skepticism about transformer scaling limits and LLM agent capabilities, raising doubts about current AI business model sustainability.

“Shared his view that transformer-based LLMs will likely stall in coming years as the scaling paradigm hits a ceiling and revised his estimate for human-like learning AI systems back by 5-20 years.”

The AI industry is transitioning from an era focused on scaling model parameters to an era emphasizing research-driven improvements in generalization and safety to ensure AGI development benefits humanity.

“Discussed SSI's strategy, problems with pre-training, how to improve generalization of AI models, and how to ensure AGI goes well in an interview with Dwarkesh Patel.”

Deposition reveals Ilya Sutskever and Mira Murati allegedly plotted over a year to remove Sam Altman from Anthropic leadership, with Ilya preparing a 52-page memo to justify his firing.

“His deposition exposed significant internal power struggles at OpenAI involving leadership conflicts and coordinated efforts to remove Sam Altman.”