Box Inc. (NYSE:BOX) CEO Aaron Levie warned that rapidly advancing AI capabilities could reshape knowledge-based professions, as agents increasingly learn to operate software, generate code and perform complex tasks across industries.
AI Agents Could Automate More Than Coding Jobs
On Sunday, in a post on X, Levie argued that AI models will continue improving at coding, mathematics, tool use and computer operations, allowing agents to take on increasingly complex professional responsibilities.
“More and more work just becomes coding-shaped to adapt to what agents are good at,” Levie wrote, suggesting that businesses could restructure workflows around tasks AI agents can perform effectively.
He pointed to recent AI developments that allow agents to generate animated videos, build games and create 3D product models using code.
Levie also highlighted potential applications in cybersecurity, life sciences and finance, arguing that AI does not need to master every subject to deliver meaningful results.
“Our mental model traditionally has been that the models have to be able to know everything about anything to generate value,” he wrote.
Instead, he argued that combining software proficiency with specialized industry knowledge could allow AI agents to make decisions and execute tasks across complex professional workflows.
“When you have that powerful combination, almost everything else gets unlocked,” Levie said, predicting that the automation of knowledge work would accelerate.
AI Agents Transforming Workflows
Last month, Microsoft Corp. (NASDAQ:MSFT) and Alphabet Inc. (NASDAQ:GOOG) (NASDAQ:GOOGL) expanded their AI agent capabilities, focusing on enterprise adoption, security and new monetization opportunities.
Microsoft CEO Satya Nadella said Copilot customers consuming usage-based credits grew more than 75% during the quarter, while its Agent 365 registry surpassed 50 million agents.
Google CEO Sundar Pichai outlined plans to integrate AI agents across Search, Gemini, Chrome, Android and Workspace.
Pichai also revealed that he personally uses AI agents to track company developments, monitor computing resources and summarize internal discussions, saving time on routine tasks.
He highlighted their ability to operate continuously, retain context and resume tasks while humans remain involved in important decisions.
Meanwhile, former Tesla AI chief and OpenAI researcher Andrej Karpathy said AI agents had transformed his programming workflow, revealing he had barely written code since December as AI increasingly handled software development.
He also discussed building Dobby, an AI-powered assistant that manages smart-home functions through natural language commands.
Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.
Photo Courtesy: Stock-Asso on Shutterstock
© 2026 Benzinga.com. Benzinga does not provide investment advice. All rights reserved.



