Replit CEO Amjad Masad has argued that the economy does not need true artificial general intelligence (AGI) to undergo a profound transformation—what he calls "functional AGI" is already sufficient and within reach. Speaking on the a16z podcast, Masad emphasized that while the idea of a human-like superintelligence captures public imagination, the real economic impact is being driven by AI systems that can learn from real-world data and autonomously complete specific, verifiable tasks. Masad defined functional AGI as AI that doesn’t require consciousness or human-level reasoning, but instead excels at adapting and performing across a wide range of practical applications. He believes such systems are already capable of automating large portions of labor across industries, making them powerful enough to reshape society and the economy. He expressed skepticism about the likelihood of achieving true AGI—the kind of AI that can understand, learn, and apply knowledge across any domain like a human mind. While acknowledging that true AGI could represent a leap forward for civilization, Masad said he is "bearish on true AGI breakthrough" because the current AI systems, despite their limitations, are already highly useful and economically valuable. He warned that the AI industry may be trapped in a "local maximum trap," where companies focus on incremental improvements to existing large language models rather than pursuing fundamentally new approaches. By optimizing for short-term gains and profitability, the field risks missing the path to a true breakthrough in general intelligence. Masad also questioned whether solving the problem of general intelligence will happen within our lifetimes, saying, "Maybe the general problem is actually not within our lifetimes. Who knows?" His views come amid growing skepticism about the feasibility of AGI. While major AI labs like OpenAI, Google, Meta, and Microsoft continue to pursue it, some experts argue that current models are not on a trajectory to achieve true general intelligence. Gary Marcus, a prominent AI researcher, has stated that pure scaling of data and compute will not lead to AGI, calling the idea of "AGI in 2027" more marketing than reality. Even OpenAI’s CEO Sam Altman admitted that GPT-5, while highly capable, still falls short of the full definition of AGI. Meta’s Yann LeCun echoed this, suggesting that AGI may still be decades away, especially since many of the most complex problems in AI do not scale well with more data or compute. Masad did not respond to a request for comment from Business Insider. Nonetheless, his perspective highlights a growing shift in thinking: the most immediate and transformative impact of AI may not come from building a superintelligent machine, but from deploying systems that are smart enough to do real work.