UnitedHealth Group just announced they're pouring $3 billion into AI across their entire operation. And honestly, the scale here is what got my attention.
They already have 22,000 software engineers on payroll. More than 80 percent of them are using AI to write code or build new agents right now. Not in a pilot program. Not in some innovation lab. In production, across the whole company.
This is the largest health insurer in the country deciding that AI isn't a side project anymore. It's the main project. They're hiring data scientists and AI specialists like they're going out of style. When a company that touches millions of patients goes this hard on automation, you better believe it's going to change how your doctor's visit works, how your claims get processed, and how fast you get a prior auth back.
My take: healthcare has been stuck in the digital dark ages for decades. If AI is what finally drags it into the modern era, I'm cautiously optimistic. But "cautiously" is doing a lot of work in that sentence. UnitedHealth doesn't exactly have a track record of putting patients first, so I'll be watching how this plays out real close.
OpenAI Wants the Government to Prepare for an AI Economy
OpenAI dropped a 13-page policy paper called "Industrial Policy for the Intelligence Age" and it reads like someone finally said the quiet part out loud.
The proposals are wild. Taxes on automated labor to replace payroll tax revenue that's about to disappear. A national public wealth fund that gives every American a stake in AI-driven growth. Pilot programs for 32-hour work weeks at full pay. Automatic safety net triggers that kick in when AI job displacement hits certain thresholds.
Read that list again. This is OpenAI, the company building the tools that will cause the displacement, telling the government to get ready for it. That's either remarkably self-aware or the most calculated PR move of the year.
My take: I don't care what you think about OpenAI's motives. The proposals themselves are worth talking about. We're going to need answers to these questions sooner than most politicians want to admit. Whether it's OpenAI's version or somebody else's, some kind of economic framework for an AI-heavy labor market is coming. Better to build it on purpose than let it happen by accident.
AI Virtual Try-On Tech Is Finally Good Enough to Matter
CNBC ran a piece on AI startups building virtual try-on technology for retail, and the headline buried the real story. Returns are one of the biggest margin killers in e-commerce. We're talking billions of dollars a year in lost revenue because people order three sizes and send two back.
Generative AI has gotten good enough that these virtual try-on tools actually work now. Not the janky overlays from five years ago. Real, convincing visualizations of how clothes fit on your specific body type. Retailers are starting to see measurable drops in return rates.
My take: this is the kind of AI application I love. Not some world-changing superintelligence pitch. Just a straightforward problem, a better solution, and money saved on both sides. The retailers keep more margin, customers stop playing the returns game, and fewer trucks drive back and forth carrying clothes nobody wanted. Sometimes the boring use cases are the best ones.