Dario Amodei, CEO of AI company Anthropic, has described the current era of human-AI collaboration as the "centaur phase" of software engineering. Drawing from the mythological centaur—a hybrid of human and horse—he explained that this stage represents a powerful partnership where humans and AI work together to achieve outcomes far beyond what either could accomplish alone. In an episode of the "Interesting Times with Ross Douthat" podcast published Thursday, Amodei used chess as a key example. Around 15 to 20 years ago, a human expert reviewing an AI’s moves could outperform both a standalone human and a standalone AI. Today, AI systems can defeat even the strongest human players without human oversight, signaling a shift in the balance of power. Amodei argued that the same evolution is now underway in software engineering. "We're already in our centaur phase for software," he said. While this phase may initially increase demand for software engineers—since teams leverage AI to build faster and better—Amodei warned it could be short-lived. He expressed concern over the disruptive impact AI will have on entry-level white-collar jobs, particularly in fields like law, finance, and consulting. He emphasized that this transformation is happening at an unprecedented pace. Unlike past shifts—from farming to factory work to knowledge-based economies, which unfolded over decades or centuries—this one is unfolding in just a few years. "This is happening over low single-digit numbers of years," he said. Amodei is among the most prominent voices cautioning that AI could automate up to half of entry-level professional roles within the next one to five years. His warnings echo those from other top AI leaders: Mustafa Suleyman of Inflection AI and Demis Hassabis of Google DeepMind have also predicted that advanced AI will automate many service-sector jobs within the next 18 months. However, not all tech executives agree. Some argue that AI will amplify human productivity rather than replace it. GitHub CEO Thomas Dohmke, speaking on a July podcast, dismissed the idea that AI could enable anyone with no coding skills to build a billion-dollar business. "The idea that AI without any coding skills lets you just build a billion-dollar business is mistaken," he said. Atlassian CEO Mike Cannon-Brookes echoed this sentiment, noting that as AI advances, people will continue to generate new ideas for technology they want to create. "Five years from now, we'll have more engineers working for our company than we do today," he said in an October interview. "They will be more efficient, but technology creation is not output-bound."