“We used AI to…”
“Our company is investing heavily in AI…”
“AI isn’t really delivering the results we’d hoped for…”
Which AI are you talking about? Because AI has become a generic term for “large language models created by either Anthropic or OpenAI.“
Most East Coast computer scientists will tell you that artificial intelligence is so much more than what you’re reading and hearing about on social media and from news agencies. And understanding the differences between AI and what they’re specifically referencing really matters.
There is only one type of AI constantly generating doomer headlines about the apocalypse and mass human extinction: the LLMs coming out of the two aforementioned companies.
No one is worried that the AI powering Waymo’s self-driving features will cause havoc, make insane decisions, and start trying to massacre crowds in cities.
No one is worried that the AI powering Deep Blue—the chess computer that beat Gary Kasparov—will suddenly come to life, gain control of nuclear weapons, and hold us all hostage.
The AI we’re all talking about is the only one with this kind of news behind it, and it’s the only one that most of us are able to engage with on a daily basis.1
But it’s not “AI“—at least, it isn’t all that artificial intelligence is. It’s one singular type of AI that can be useful for a few really specific things, and pretty shitty at everything else.
We need to be really clear about what we’re talking about when we say things like, “AI will destroy the planet,” or “AI will take all of our jobs away.” Because it’s not true.
AI is not doing these things. AI is not breaking out of sandboxes and causing security nightmares for businesses. It’s LLM-powered querying agents with no human oversight and no guardrails, simply acting in accordance with their programming. And that distinction matters more than you might realize.
So, when someone is ranting and raving to you about AI, ask them, “What are you talking about specifically?” It will help us all change this conversation for the better.
- This AI, by the way, is being run by crazy Rationalist/Effective Altruist Evangelicals who believe it’s their responsibility to bring forth an AI deity to save humanity—but only the few thousand of us (actually, just them, the rich billionaires) that they can upload into the digital aether to surf the stars (seriously).
And what’s better? All of this is based on a blog from the 2000s written by an 8th-grade dropout with no computer science or engineering background and financed by Peter Thiel, who is currently ranting about the anti-Christ and claiming, with seemingly no understanding of irony at all, that “evil is good.” ↩︎
