Zach Wills outlines his predictions and priorities for AI in 2026, emphasizing a shift towards more complex, agent-based systems. He believes that multi-step, self-correcting loops will become standard in AI workflows, surpassing simpler one-shot prompts. Companies that can create AI-native products—those that replace entire workflows rather than just enhancing them—will see higher valuations. Wills cites a Bessemer report showing that AI-native applications command greater revenue per employee and deliver complete outcomes, which is key for standing out in the market. Wills also highlights the importance of building internal tools that utilize AI agents to automate tedious tasks, which can drive efficiency and enhance service quality. He notes that the cost of development has decreased, making it feasible for companies to create tools previously deemed too complex. The gap between those who effectively adopt these technologies and laggards will become stark, leading to a divide among individual contributors. Wills suggests that the coming year will reveal a clear distinction between AI-native superstars—who craft their own sophisticated workflows—and those still relying on traditional methods.