Anthropic named their AI after a guy nobody's heard of. He beat 1,025 out of 1,026 fund managers from his kitchen using an Apple 2...
His name was Claude Shannon.
In 1948 he wrote a paper. It proved that a photo, a song, a phone call and a letter are all the same thing underneath...
Just ones and zeros.
He called the unit a "bit"
Everything digital you use sits on top of that btw...
Your payments. Your video calls. Your ads. The AI you use every day.
Anthropic named their model Claude after him.
He also made 28% a year in the stock market for thirty years as a hobby.
He did it from his kitchen with his wife and an old Apple computer lol.
In 1986 a magazine ranked 1,026 funds...
He beat 1,025 of them.
Here's how he said he solved problems:
- Cut it down
He said most problems are buried in extra info. Take everything out except the part that actually matters. I tried this on my churn. Wrote down 11 reasons clients leave. 9 of them were things I tracked because they were easy to track. 2 actually mattered. I'd spent 3 months on the 9.
- Look at problems already solved
Don't attack yours head on. Find every similar problem that's already been figured out. Then look at what those answers have in common. He said two small jumps are much easier than one big jump.
- Say the question a different way
Change the words. Change the angle. He warned about sunk cost here. The work you've already done is usually the thing keeping you stuck. That's why someone brand new can crack it in ten minutes.
- Break it up and let yourself wander
A lot of big discoveries came from someone poking at unrelated stuff that went nowhere. Then hitting the answer from the side.
- Flip it
Can't get from here to the answer? Assume the answer is already true. Then work backwards. So instead of "how do I get to 100k a month," assume you're already there. What are you selling? To who? How many? What breaks at that size? Totally different questions. And you can actually answer all of them.
- Make it bigger
Once you solve one version, build the one that solves all of them. You fixed one client's onboarding. Now make the version that works for every client. That's the difference between a job and an asset.
Here's how he labels someone that can solve problems quickly:
"A genius is simply someone who is usefully irritated."
That itch when something isn't right and you can't leave it alone.
He called it constructive dissatisfaction.
Two more things...
His mentor's rule was that specialising kills genius.
So he'd throw Shannon into fields he knew nothing about on purpose.
And Shannon worked on three unrelated problems at once...
Said it beat focusing on one (which can be kinda backwards to what people say today)
P.S - Book is " a mind at play" by Jimmy Soni and Rob Goodman