Elon Musk got rejected by Netscape. He walked into the lobby, was too shy to talk to anyone, and walked out. Never got the job.
At his first company Zip2, the board demoted him. Twice. They refused to let him be CEO.
He got fired from PayPal as CEO while flying to his own honeymoon. The board voted him out mid air.
He almost died of malaria in 2000. Ten days in intensive care. Lost 45 pounds. A day from death.
His first child died at 10 weeks old.
His first rocket exploded. Falcon 1, flight one. Burned on the pad.
His second rocket exploded.
His third rocket exploded. The last of his money was nearly gone.
Tesla nearly went bankrupt in 2008. The closest he ever came to a nervous breakdown.
Both companies almost died on the same Christmas Eve.
He was sued by investors. Mocked by the people who built cars before him.
His childhood heroes, the astronauts who inspired him, testified against his company to Congress.
The Cybertruck window shattered on live stage in front of the world.
He overpaid for Twitter by his own admission and watched its value collapse.
He was beaten unconscious as a child and thrown down a flight of stairs.
He has said he goes to sleep alone and it kills him.
He failed in public, over and over, for thirty years.
He is the richest man in the history of the world.
The difference was never the absence of failure. It was the refusal to stop after it.
In the AI sector, the AI coding space saw explosive, hockey-stick growth from early last year through June of this year. Anthropic perfectly timed the wave, scaling its ARR in just one year to surpass OpenAI and claim leadership of the category.
However, this phase is likely to moderate in the second half of the year. The core reason is that AI token costs have yet to see a true step-function breakthrough. Real inflection will only arrive when token economics plummet—much like internet access costs did decades ago—turning AI into a ubiquitous, must-have foundation for everyone.
That's when the second act begins: the explosive takeoff of AI agent "vibe coding," especially autonomous agents. Only then will a new generation of giants emerge to surpass today's leaders.
Bottom line: The past 18 months of AI coding demand has been primarily for programmers and developers. The next massive leg up will be for agents—and ultimately for everyday people.
In the near future, today's mathematicians and scientists will be viewed the same way we now look back on ancient wizards from thousands of years ago, Confucian scholars from centuries past, and the priests of old churches — fascinating relics of a pre-AI era.
In the coming Algorithmic Paradigm Era, we will face an inevitable explosion of mathematical proofs. The vast majority will surpass humanity's limited time, energy, and cognitive bandwidth to fully understand, analyze, or explain. So how do we evaluate and incentivize these formally verified, correct mathematical proofs that humans can no longer comprehend?
VibeWant's Human- or AI-Created Knowledge Valuation Framework offers a powerful new solution. For novel, formally verified mathematical proofs and the emergent knowledge generated by AI, there is no need for human authorities or gatekeepers. Instead, value is discovered, priced, and rewarded through pure market mechanisms. Over time, the laws of nature and evolutionary dynamics will naturally filter and surface the most valuable knowledge for both the global AI network and human society.