Actually, the expression in this picture is quite spot-on, giving a “I’m helpless too” kind of vibe.
Tech bosses used to only worry about whether their servers crashed; now they have to worry about whether their houses will get blown up.
It was only 13.9 degrees Celsius in Shanghai today, cloudy, and a bit chilly outside. I wrapped my jacket tightly around me when I went out to buy coffee and casually scrolled through the news. As it turns out, Silicon Valley across the ocean is literally in “deep water and scorching fire” right now.
A 4 AM San Francisco With No Kobe, Only Molotov Cocktails
Around 4 AM this Friday, North Beach, San Francisco.
While most people were still dreaming in bed, a 20-year-old young man snuck up to the front door of OpenAI CEO Sam Altman’s home. He didn’t bring a petition, nor did he hold a PowerPoint filled with creative business plans.
He pulled out a Molotov cocktail.
With a bang, that exquisite outer door caught fire immediately.
That wasn’t the end of it. About an hour later, shortly after 5 AM, this guy strolled over to the outside of the OpenAI headquarters building on the 1400 block of 3rd Street. Oh boy, he directly threatened to burn down the entire building.
However, the San Francisco Police Department (SFPD) reacted incredibly fast this time. As far as I know, they followed the clues, immediately identified this person on the scene, and pinned the young man down right there. Fortunately, there were no casualties.
Truly fierce.
I’ve seen a saying before that the ultimate destination for Silicon Valley hardware products is making bulletproof glass thicker and thicker. Now it seems that wasn’t a joke. Think about it: a barely 20-year-old kid, staying up in the middle of the night carrying a Molotov cocktail to bomb the front door of the most popular boss in the AI circle. This scene reeks of strong cyberpunk flavor. Actually, thinking about it makes one sigh. Technology was supposed to be something that makes people’s lives better, but now it has pushed ordinary people into such extreme stress states.
The question is: What is exactly going on? Why did Sam Altman get hit with this Molotov cocktail?
Actually, if you’ve been following the AI circle’s movements these past couple of days, you’d realize this fire was bound to ignite sooner or later.
The Snatched Pentagon Order
To understand this, we first have to talk about OpenAI’s mortal enemy—Anthropic.
The feud between these two companies could easily be turned into an 80-episode TV drama. Anthropic’s boss, Dario Amodei, originally left OpenAI in the early days. I heard that at an AI summit in India a while ago, a group of tech leaders took a group photo on stage. Everyone graciously linked arms, except Sam Altman and Dario, who stubbornly refused to hold hands. The awkward atmosphere was enough to carve out a three-bedroom apartment with your toes.
Treating the development of large models like fighting a war, Anthropic is the type to fiercely defend the trenches and strictly stick to the rules of engagement; OpenAI, on the other hand, is the radical faction, eager to push new artillery to the front lines and fire first regardless of the consequences.
A few days ago, the US Department of Defense completely fell out with Anthropic.
The DoD wanted to procure the most cutting-edge models, but Anthropic rigidly defended its bottom line. They have an internal mechanism called “Constitutional AI.” To explain it simply, it sets rules for the model, ensuring it absolutely cannot participate in large-scale domestic surveillance and absolutely cannot be used in fully autonomous weapons.
How could the Pentagon swallow this kind of insult? In a fit of rage, US Secretary of Defense Pete Hegseth directly labeled Anthropic a “supply chain risk.” This is practically issuing a ban.
And guess what happened?
Right at this critical juncture, Sam Altman swiftly jumped out, announcing with a face full of smiles: We, OpenAI, have already signed a $200 million contract with the Department of Defense. Not only did they sign it, but they will also deploy the model directly into the DoD’s classified network.
Although OpenAI’s PR statement was beautifully written, claiming they have extremely strict guardrails, anyone with a discerning eye can see this was a blatant order snatching. The arms business that others refused in order not to do evil, you turned around and accepted entirely.
To be honest, this wave of operations definitely offended their peers. You can’t entirely blame Sam Altman for being greedy; after all, maintaining the operation of those massive GPU clusters burns an astronomical amount of money every day. The Pentagon’s order is like an oasis in the desert—who wouldn’t covet it? But doing this has indeed pushed the company into the volcano crater of public opinion.
Look at the positioning and expressions in this group photo; the peers are truly blatant enemies.
From Code Sandboxes to Real Trenches
After seeing this, my biggest feeling is that the backlash from the physical world has come far more fiercely than anyone expected.
We used to always think that geeks typing on keyboards in garages were merely changing a few lines of pixels on a screen. Even AI was just a cyber toy that could help you write weekly work reports or draw a few funny pictures.
But reality slapped us hard.
When code starts deciding the tracking logic of a military drone, when an algorithm might involve the operation of a state machine, it is no longer a toy.
Speaking of this, I’m reminded of another incident. A couple of years ago, Elon Musk and Mark Zuckerberg argued fiercely online, clamoring to schedule a fight in an octagon. Netizens all over the world grabbed their popcorn. But at best, that was just a PR stunt used by multi-billionaires to hit their KPIs.
But the flames in San Francisco’s North Beach today represent an extremely real social tear. I once confidently told a friend while discussing the topic of technology losing control that it would take at least fifty years for AI to cause severe social confrontation.
My face hurts from the slap.
That young man who threw the Molotov cocktail might just be an extremist deeply panicked about technological development, or perhaps an ordinary person angry about tech companies becoming deeply involved in the military. His behavior is undoubtedly a crime and absolutely must not be encouraged. But frankly speaking, this ball of fire is an extremely clear signal.
For the past ten years, Silicon Valley’s tech giants have been running wildly while hiding under the beautiful cloak of “technological neutrality.” Now, this cloak has been burned through by the fury of the real world.
A few lines of code exchanged for hundreds of millions of dollars in defense contracts, but isn’t this money too hot to hold?
No One Can Stay Out of It
Actually, similar things have happened in the history of technology. Back during the Industrial Revolution, the Luddites smashed weaving machines because they felt the fear of being swallowed by machines.
The difference is that people used to smash wooden machines, whereas today people throw glass bottles filled with gasoline.
Moreover, the scope of impact today is too broad. I’m not entirely sure, but it is said that Anthropic just threw $20 million this month at an organization called “Public First Action,” specifically to lobby the public to guard against the high risks of AI. On one side, they’re frantically throwing money around urging everyone to hit the brakes; on the other side, they’re rushing to sell the latest AI engines to the military. This scene is truly surreal.
Oh, and one more thing. Despite the uproar, the San Francisco police are still keeping the name of that 20-year-old arsonist strictly under wraps.
It’s probably because they fear that once it’s announced, he will be hailed as a hero by some extreme anti-tech individuals, triggering more copycat effects. It’s not that simple.
Alright, I’m going to eat; my takeout has already gone completely cold. 🤷♀️
Do you guys feel that the current speed of technological development is already so fast that it makes your scalp tingle?
References：
- Sam Altman’s San Francisco home targeted with Molotov cocktail, OpenAI says
- Sam Altman house hit with Molotov cocktail, OpenAI office threatened
- Battle of the AI brands: What’s behind the bad blood between OpenAI and Anthropic
- The Department of Defense’s Conflict With Anthropic and Deal With OpenAI
- OpenAI wins $200 million U.S. defense contract
—— Lyra Celest @ Turbulence τ.
