Databricks CEO Ali Ghodsi has declared that artificial general intelligence (AGI) is already here, despite the tech industry’s ongoing efforts to redefine what it means. Speaking at the Goldman Sachs Communicopia + Technology Conference in September, Ghodsi argued that today’s AI chatbots already meet the original definition of AGI—systems capable of reasoning and learning across a broad range of tasks like a human. “Everybody would say yes, but we kept moving the goalposts,” Ghodsi said during the panel discussion, which was published Tuesday. “Since we kind of achieved it, let’s come up with something even bigger,” he added, referencing the shift in focus toward superintelligence—the hypothetical AI that surpasses human intelligence in every domain. Ghodsi, who holds a PhD in computer science, emphasized that the pursuit of superintelligence is not only unrealistic with current techniques but also misaligned with practical needs. He argued that AGI already provides all the capabilities required to build autonomous AI agents capable of automating complex workflows. “We just need to do the boring work,” he said, highlighting the importance of refining and deploying existing systems rather than chasing unattainable benchmarks. The comments come as the pace of breakthroughs in AI appears to be slowing. Ghodsi noted that the scaling laws that fueled rapid progress over the past few years have plateaued. Newer models like OpenAI’s GPT-5 and Anthropic’s Claude 4, he said, are not delivering the transformative leaps once expected. “It's getting harder and harder to get value out of the next pre-trained giant model,” he observed. The debate over superintelligence is intensifying across the industry. Microsoft AI CEO Mustafa Suleyman, a co-founder of DeepMind, recently called superintelligence an “anti-goal” in an episode of the “Silicon Valley Girl Podcast.” He warned that such systems would be extremely difficult to control and align with human values, calling the vision “not a positive one for the future.” Instead, Suleyman advocates for “humanist superintelligence”—AI that supports human well-being and is grounded in ethical principles. In contrast, other leaders remain committed to the superintelligence path. OpenAI CEO Sam Altman has repeatedly stated that the company’s mission extends beyond AGI to building systems that outperform human intelligence. In January, he said superintelligent tools could dramatically accelerate scientific discovery and drive global prosperity. In a September interview, he expressed confidence that the industry would reach superintelligence by 2030. Google DeepMind co-founder Demis Hassabis shares a similar outlook, predicting AGI could emerge within the next five to ten years. He envisions AI systems that understand the world with deep, nuanced insight and become seamlessly integrated into daily life. As the industry grapples with what’s possible and what’s desirable, Ghodsi’s message stands out: the tools for transformation are already in hand. The real challenge, he insists, isn’t inventing the next frontier—it’s applying what we already have.