The report also confirms what a lot of us already felt in our gut. The top models keep getting better despite all the "we're hitting a wall" talk. And AI companies are generating revenue faster than any previous tech wave, even as they burn hundreds of billions on data centers and chips. The money is real. The growth is real. The question is whether the spend catches up before the returns do.
My take: if you're still sitting on the sidelines thinking this is overhyped, you're running out of runway. The Stanford data doesn't lie. This thing is moving faster than anything we've seen, and the gap between early adopters and everyone else is getting wider by the month.
4Chan Gamers Beat Google to Chain of Thought Reasoning
This one made me laugh. Turns out a bunch of folks on 4chan playing AI Dungeon accidentally discovered chain of thought reasoning back in 2022. That's more than a year before Google researchers published their paper claiming to be the first to pull it off with large language models.
The gamers figured out that if you told the AI characters to solve math problems step by step, the answers got dramatically better. They weren't trying to advance AI research. They were just trying to get their text adventure game to stop being dumb at arithmetic.
My take: this is a perfect reminder that breakthroughs don't always come from labs with billion dollar budgets. Sometimes they come from people just messing around and paying attention to what works. The history of AI is going to have some wild footnotes, and "4chan gamers did it first" might be one of the best.
One Son, One AI Workflow, Three Saved Interventions
Pratik Desai is a 34 year old technologist whose mother was diagnosed with Stage 4 duodenal adenocarcinoma. Instead of just trusting the system to catch everything, he built an AI workflow that ingested her daily medical records from Epic into NotebookLM and Claude.
The result? He spotted a CAT scan misdiagnosis, caught a medical emergency the care team missed, and coordinated interventions that likely extended her life. Three critical catches. Not from a fancy hospital AI system. From a son with consumer tools and determination.
My take: this is the story that matters most today. Not because AI replaced doctors, but because it gave a regular person the ability to be a better advocate. Healthcare is drowning in data and complexity. AI tools that help families navigate that mess aren't a luxury anymore. They're becoming a necessity. If you have a loved one dealing with a serious diagnosis, start looking into how these tools can help you ask better questions and catch what falls through the cracks.