AI agents are moving into real work. That could change how software companies get paid (www.cnbc.com)

AI agents are transitioning from theoretical applications to practical implementations in business workflows, particularly in customer service, sales, and support. Clay Bavor, co-founder of Sierra, discusses this shift in an interview, highlighting how organizations are increasingly focused on deploying AI agents that not only respond to inquiries but also perform complex tasks. This evolving landscape is prompting companies to seek clearer metrics for assessing the return on investment (ROI) from these technologies. The significance of this development lies in the potential shift towards outcome-based pricing models, which could fundamentally alter the revenue structures for software companies. Instead of traditional licensing fees, businesses may pay based on the specific outcomes generated by AI agents. Bavor also notes challenges associated with deploying these agents, referring to the "last mile" of implementation as a critical hurdle. Additional topics in the discussion include the rising costs of AI tokens and the emergence of coding agents, which further emphasize the accelerating integration of AI into practical business functions.