Databricks CEO says AGI is already here — and Silicon Valley just keeps moving the goalposts

In a bold assertion at the Goldman Sachs Communicopia + Technology Conference, Databricks CEO Ali Ghodsi claimed that artificial general intelligence (AGI) is already a reality, despite Silicon Valley's ongoing hesitation to accept this fact. Ghodsi remarked that current AI chatbots fulfill the criteria for AGI—defined as AI capable of human-like reasoning—according to standards established a decade ago. "Everyone would agree, but we keep shifting the goalposts," Ghodsi stated during the discussion, which was made public recently. He pointed out that instead of recognizing this achievement, the focus has transitioned to a more ambitious goal: superintelligence, which would surpass human reasoning capabilities. Ghodsi expressed concern that the industry's obsession with superintelligence is misguided, arguing that businesses should prioritize creating systems that can automate tasks effectively rather than striving to outsmart the best human intellects. He emphasized that AGI already possesses the necessary capabilities for automation and developing intelligent agents, adding, "We just need to focus on the routine tasks." Databricks, headquartered in San Francisco, recently secured $1 billion in funding, bringing its valuation to over $100 billion. During the conference, Ghodsi noted that the rapid advancements in AI models appear to have plateaued. He explained that the scaling laws that have driven AI progress in recent years have reached a halt, with newer iterations such as OpenAI's GPT-5 and Anthropic's Claude 4 failing to deliver significant enhancements. "Extracting value from the next large pre-trained model is becoming increasingly challenging," he remarked. Ghodsi's comments emerge amid a divide in the tech sector regarding the pursuit of artificial superintelligence. Mustafa Suleyman, CEO of Microsoft AI, recently labeled superintelligence as an "anti-goal" on the "Silicon Valley Girl Podcast." He expressed skepticism about the positive implications of such an advancement, underscoring the difficulty of aligning superintelligence with human values. Conversely, other industry leaders remain committed to the quest for superintelligence. OpenAI's CEO, Sam Altman, has stated that the company is focused on this goal rather than stopping at AGI, envisioning superintelligent tools that could revolutionize scientific discovery and innovation. Altman speculated in a September interview that he would be surprised if the industry had not achieved superintelligence by 2030. Similarly, Demis Hassabis, co-founder of Google DeepMind, projected a timeline for AGI's emergence within the next five to ten years, envisioning a future where AI systems possess a profound and nuanced understanding of the world, seamlessly integrated into daily life.