David J. Blumberg

57% of white-collar work hours could be automated within a decade. The contrarian part: that may increase employment, not reduce it. Most people misunderstand what AI actually does. AI does not primarily eliminate work. It eliminates friction. The industrial revolution did not end labor. It massively increased productivity, lowered costs, expanded demand, and created entirely new industries. AI agents may do the same thing to knowledge work. One example from my conversation with @davidblumberg: A medical device manufacturer increased production from 10,000 surgical rings per day to 166,000 while improving defect detection from 94% to 100%. Same industry. Same product. Completely different productivity curve. The biggest beneficiaries of AI may not be consumers. They may be the workers and companies that learn to compound output faster than competitors. One line stood out to me: “you’re more likely to lose a job not to an AI, but to someone else who is using an AI when you’re not.” That feels directionally right. The market is still treating AI like a software upgrade. I increasingly think it is a labor force multiplier. And historically, labor force multipliers tend to expand economies, not shrink them. Curious if executives or allocators reading this disagree: does AI ultimately compress labor demand, or does it create a new productivity supercycle that increases it? We’d like to thank @AlphaSenseInc for sponsoring this episode! #ArtificialIntelligence #AgenticAI #VentureCapital #B2BSoftware #EnterpriseAI #FutureOfWork Full episode below 👇 youtube.com/watch?v=Tly916EQ…

The entire software industry has spent 30 years fighting over $10,000 per employee. The real prize is the other $160,000. @davidblumberg, founder of Blumberg Capital, ran the numbers on what agentic AI is actually coming for: "The average white collar worker costs the employer about $170,000. $160,000 is wages and benefits. $10,000 is spent on software tools." Microsoft, Google, Oracle, AWS, a $17 trillion industry, have been competing for that $10K slice. Agentic AI goes after the $160K. "McKinsey says agentic AI, within a decade, is likely to transform 57% of those work hours." That is roughly $100,000 of wages per knowledge worker moving from human payroll to software. Which is why Blumberg's conclusion is so aggressive: "The agentic agent software market will be two or three times larger than the existing software industry today." The biggest software companies ever built were priced off the $10K. The next ones get priced off the $160K.