Anthropic CEO Predicts Human-Level AI Could Arrive as Soon as 2026

Anthropic CEO Dario Amodei has suggested that artificial intelligence capable of performing at human levels may emerge by 2026, should current advancement rates hold steady. Speaking with podcaster Lex Fridman, Amodei likened recent progress in artificial general intelligence (AGI) to progressing through educational stages: "We’re starting to get to PhD level, and last year we were at the undergraduate level,” he explained, predicting that AI could exceed human capabilities within just a few years.

“If you just eyeball the rate at which these capabilities are increasing, it does make you think that we’ll get there by 2026 or 2027,” he noted, though he acknowledged possible setbacks, such as data shortages, limitations in scaling AI clusters, and potential geopolitical issues impacting microchip supply chains.

Factors That Could Affect the Timeline

Amodei emphasized that while progress in AI has been rapid, several obstacles could delay AGI timelines:

• Data Limitations: As AI models require increasingly larger datasets, the risk of running out of accessible, high-quality data may hinder progress.

• Scaling Challenges: Current AI models need extensive computational resources, and the ability to scale these resources may eventually reach physical or financial limits.

• Geopolitical Factors: Supply chain disruptions—especially related to microchips essential for high-performance computing—could slow down development.

Despite these challenges, Amodei believes that “if you believe the straight line extrapolation,” human-level AI could be reached by 2026 or 2027, though minor delays remain possible.

Balancing Advancements with Responsibility

Amodei voiced concerns about the profound impact of human-level AI, noting, “Things that are powerful can do good things, and they can do bad things.” He emphasized the ethical responsibility involved, highlighting that “with great power comes great responsibility.” Anthropic’s mission, he explained, is to drive progress responsibly, focusing on setting a positive example for other AI developers.

Anthropic’s main product, the large language model Claude, competes with similar models like OpenAI’s ChatGPT. Amodei reaffirmed that Anthropic aims to “push the other players to do the right thing by setting an example,” underscoring the company’s goal of promoting safe AI practices.

Competition in the AGI Race

AGI is generally defined as AI that matches or surpasses human abilities across a wide range of cognitive tasks, presenting significant implications for industries, governments, and society.

OpenAI’s CEO Sam Altman has made similar forecasts for AGI, predicting its arrival as early as 2025. Altman outlined OpenAI’s strategy to accelerate AGI by refining infrastructure, aligning products, and advancing research. To illustrate the path toward AGI, Altman shared a framework with five levels of AI development:

• Level 1: Basic chatbots capable of generating responses and engaging in simple conversations.

• Level 2: Systems with reasoning abilities, which OpenAI believes it reached this year with advanced capabilities in reasoning and problem-solving through its o1 preview model.

• Level 3: Agent-based AI that can independently carry out tasks and interact over time, asking for help or adjusting its approach when needed.

• Level 4: Systems that function as innovators or scientists, able to explore unknown areas and solve complex problems without constant human guidance.

• Level 5: An organizational-level intelligence that can perform tasks at the scale of an entire company or complex entity.

OpenAI is currently advancing toward Level 3, and Altman believes that innovation and creativity within today’s models may bring these levels faster than anticipated.

Preparing for the Implications of Human-Level AI

As the AI industry races toward AGI, business leaders and policymakers are urged to consider the profound social and ethical implications. Companies need to prepare for the regulatory landscape, anticipate challenges in data management, and understand the potential impacts of AGI on the workforce and society.

If achieved, AGI could transform industries, automate complex tasks, and revolutionize innovation. However, stakeholders must approach its development with a commitment to ethical standards and long-term societal impact, balancing innovation with responsibility.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.