Masayoshi Son announced that AI is now designing OpenAI's next model, which he considers evidence of "superintelligence." The man who lost $17 billion on WeWork just redefined intelligence itself. Inspiring stuff.
Son told CNBC his previous forecast of superintelligence arriving in 10 years was "conservative." He thinks it'll be here sooner. This is the same executive who thought Adam Neumann selling desks was worth $47 billion. His track record on predictions sits somewhere between a Magic 8-Ball and your uncle who swears the Falcons will win the Super Bowl this year. But sure, let's trust his timeline on artificial superintelligence. The machines will definitely achieve consciousness right after SoftBank's Vision Fund posts its first profitable quarter.
OpenAI has apparently handed over model design to AI. The robots are building the robots now. Son sees this as a sign we've crossed some threshold. I see it as OpenAI's engineering team discovering they can blame the AI when the next model hallucinates even harder than GPT-4. "Sorry your chatbot told you to eat rocks. The superintelligence made an architectural decision we don't understand." Plausible deniability with a PhD.
SoftBank invested in this. SoftBank, whose greatest hit was accidentally making money on Alibaba and then spending the next two decades trying to replicate that high. Their portfolio reads like a suicide note written by an investment committee. Every company either imploded spectacularly or exists in that special purgatory where nobody remembers it launched. But Masa Son knows exactly when superintelligence arrives. He's got a gift.
Retail traders will read this headline and immediately buy $NVDA calls. They'll convince themselves they're positioned for the AI revolution. They'll imagine themselves getting rich as the machines wake up. They won't stop to consider that if AI becomes superintelligent enough to design itself, it'll probably be superintelligent enough to front-run every trade their Robinhood account attempts. The singularity arrives and the first thing it does is liquidate your portfolio.
Son's conservative estimate just became less conservative. Which means he's more bullish on robot god than he was on Uber at a $120 billion valuation, and that worked out perfectly for absolutely nobody.
Photo by on Unsplash

Leave a Comment