Peter Lynch turned $18M into $14B by simply following these 6 stock investment rules:
1. Trailing P/E < 25x
2. Forward P/E < 15x
3. EPS Growth > 15%
4. PEG Ratio < 2
5. Debt/Equity < 35%
6. Market Cap > $5B
Short, simple, & has proven success for decades.
Save this for later…
This is probably the best tweet you will read today.
A kid who was bad at maths, rejected by 25 universities and turned down by DeepMind three times, went on to build the software that OpenAI, Tesla and Meta now use to train their AI.
His story is the best argument I've seen for never ever quitting.
He grew up in Hyderabad (India), struggled with maths at school, and ended up at VIT in Vellore. VIT is a kind of college people call tier 2 in India, or probably worse.
He applied to twelve American universities for a master's with a GRE score of 1420 and was rejected by every one of them. He flew to the United States anyway on a J-1 visa with no real plan, applied to fifteen more programmes, and was turned down by all of them except USC and, through a late admission, NYU in 2010.
At NYU he started working with a researcher named Yann LeCun. Yes the same legendary researcher you’re thinking about. When he looked for jobs, the rejections kept coming. DeepMind said no. Almost everyone said no. His only offer was a test engineering role at Amazon.
Later, he applied to deepmind again, but was rejected again.
He kept building anyway. Around 2011 he wrote one of the fastest AI inference engines on phones, and still got rejected by DeepMind a third time, and then emailed LeCun one more time.
That email, along with his open source work, got him into Facebook's AI research lab, where he barely made it through bootcamp. Then senior engineers several levels above him couldn't get a major image model to train, and he, the most junior person in the room, found the bug that had everyone stuck. It was his first real win, almost ten years in.
He went on to co-create PyTorch with a team of three, and then watched management try to shut it down over internal politics. He has said he cried at a bar about it. A few people fought to save it, it launched in 2017, and today it sits underneath a huge share of the AI the world uses.
His name is Soumith Chintala.
Today he is Chief Technology Officer at Mira Murati led Thinking Machines Labs.
When the going gets tough. The tough gets going.
source - @autumn_ai_inc
Elon Musk is not actually prepared for what's going to happen if Democrats win one or both chambers of Congress after he's spent the last few years as an overtly Republican megadonor with a lot of self-dealing contracts in D.C.
He's never had to deal with a majority party that's outright hostile to him and his businesses.
Remember what a massive crashout he had just because Biden didn't invite Tesla to an electric vehicle summit one time.
Now imagine how he's going to react when there are House committees spending massive amounts of time and energy investigating his companies for public corruption.
$CRWV $ORCL $NBIS these companies have real revenues. Not sure why everyone is suddenly Elon's pawns. In fact, the neo labs will need the likes of CoreWeave and Nebius b/c SpaceX is a competitor.
@jiahanjimliu@elonmusk As of now there is 0 compute $SPCX has deployed. They will start deploying in late 2027. But yeh Mr. Musk is worried about debt market shut down for other neoclouds. This is the advantage you have when you own x , so that you can spread FUD