Copilot productivity, why agents are the next frontier and shaping the future of India with AI

A Note from Malcolm

Microsoft distinguished itself in Big Tech last week, reporting robust second-quarter financial results.  Satya Nadella pointed to Github Copilot as being central to growth as it “is by far the most widely adopted AI-powered developer tool.  Just over two years since its general availability, more than 77,000 organizations from BBVA, FedEx and H&M to Infosys and Paytm have adopted Copilot.”  

Impressive adoption, yet it begs the question: Where’s the beef?  That is, where is it producing meaningful business value?  This isn’t an isolated question.

At TalentGenius we’ve been closely monitoring the code-generation space as it can provide meaningful insight on how and where software will eat work.  More than $2 billion of venture capital has been invested across market leaders Copilot, Cursor, Replit Ghostwriter, Augment, Sourcegraph Cody, Codeium and Kite. Overall their sales have been impressive.  

Yet, to date, these code generation tools have not yielded significant business results. Overall developer productivity hasn't markedly increased, nor have workforces shrunk. Tools alone don’t drive step-change.  This mirrors historical gaps between innovation and application, like the century between Newcomen’s steam engine and Britain’s first railroads. Realizing generative AI's full potential in software development requires new approaches and methods.  The individuals and companies that figure that out in the next year will be greatly rewarded, while those who miss it may be in for a bumpy ride.  

With that lens, here are some highlights from the past week at the intersection of enterprise AI and work: