Former OpenAI AGI Lead Predicts AI's Expansion to Human-Level Computer Tasks

The realm of artificial general intelligence (AGI) remains shrouded in ambiguity. This concept refers to an advanced form of artificial intelligence potentially capable of matching or surpassing the intellectual capabilities of humans.

However, breakthroughs in technology indicate that we may soon witness a version of AGI. Leading experts suggest the technology could soon mirror many human abilities performed on computers. Miles Brundage, who formerly directed AGI policy and readiness at OpenAI, articulated in an interview on the tech podcast Hard Fork that the industry is on the cusp of building machines able to perform any task a person can do remotely on a computer. This encompasses tasks like manipulating a mouse and keyboard or even participating in video calls with a lifelike presence.

Brundage emphasized the importance for governments to consider the implications such technologies bring to sectors like taxation and education, suggesting potential areas for adaptation and investment.

The anticipation of companies such as OpenAI achieving AGI generates significant debate among enthusiasts and critics alike, with some expecting its advent within a few years. Notably, John Schulman, a cofounder of OpenAI who left in August, and Dario Amodei, who leads Anthropic, a competitor in the field, both foresee a timeline where AGI might materialize in the near future, with Amodei envisioning a version ready by 2026.

Given his extensive tenure over six years at OpenAI, Brundage is particularly well-versed in the company's plans and strategies concerning AGI.

Insights from the Company

At OpenAI, Brundage played a crucial role in providing guidance to executives and board members on AGI preparedness. His contributions were integral to pioneering safety research methods, including 'external red teaming', which involves enlisting external experts to scrutinize and identify potential issues within OpenAI's innovations.

The company has seen several prominent departures, with some individuals citing concerns over the equilibrium between AGI progression and safety protocols. Nevertheless, Brundage has stated that his exit was unrelated to these particular worries.

On the tech podcast, he expressed his confidence, noting, 'There isn't any lab that has completely mastered the landscape,' affirming his belief in the uniqueness of OpenAI's efforts.

Motivations and Future Endeavors

In his departure announcement on X, Brundage revealed his aspiration to make more impactful contributions as a policy analyst or advocate within the nonprofit arena. He emphasized on Hard Fork the continued alignment with his decision, elaborating on his departure motives.

He mentioned that he was unable to address all pertinent industry-spanning issues at OpenAI, such as the formulation of regulations. He further stated his desire for independence to prevent his perspectives from being dismissed as corporate jargon.