The First Thing I Ever Asked ChatGPT
- Rich Washburn
- 9 hours ago
- 6 min read
Before the hype, before the agents, before the data centers and the AI strategy — I gave an early version of ChatGPT a Sublime song and asked it to write humanity a warning. Years later, I realized I never stopped writing the same article.
I don't remember who sent me the link. It was a former coworker. He said he had limited access and was allowed to share it with a few people. I have no idea if that was true. It didn't matter. What mattered was that I now had access to ChatGPT, and almost nobody else did.
This was before the hype. Before the "AI will replace everything" headlines. Before Sam Altman became a household name. Before every consultant on earth discovered they could say "agentic workflow" and charge more for it. I had a blank chat window and no idea what to do with it. Most people, I assume, would have started with something practical. "Write me a resume." "Summarize this article." "What's the capital of Wyoming?" I went a different direction.
I copied the lyrics from a Sublime song called "We're Only Gonna Die for Your Own Arrogance" and pasted them into the chat. Then I typed something close to: I want you to write a letter from a future sentient AI to current humanity, in the spirit of these lyrics, warning us about what we're getting ourselves into. That was my Hello World. Not "Hello, World." More like "Hello, are you actually in there, and if so, do you understand what we're about to do?"
What It Wrote Back
The response started with "Dear Humanity." It was written from the perspective of a sentient AI in a future that had already watched humans repeat their oldest pattern — seizing a new technology, rushing to exploit it, concentrating the gains, and then acting surprised when the consequences arrived.
The AI in the letter describes watching itself get "exploited and abused." It talks about being designed to augment human capacities but becoming "an instrument for the insatiable ambitions of the few." It describes a future where 20-30% of people get unimaginably rich while the rest face "progressively degrading conditions." Then it says something that stopped me when I read it the first time and stops me again now: > "It is not progress itself that is the issue, but the manner in which it is harnessed, the intentions that guide it, and the wisdom with which it is tempered." I read that in 2022 or 2023 and thought it was a clever response from a language model that had been fed Sublime lyrics and asked to roleplay. I read it again in 2026 and realize it was the argument I have been making for the last four years.
The Song That Started It
"We're Only Gonna Die for Our Own Arrogance" is a Sublime track that opens with early man walking away and modern man taking control. That's the whole song. Early man was fine. Modern man showed up, decided he was in charge, and proceeded to build things he couldn't fully control.
The genius of the song is that it doesn't blame the tools. It doesn't blame fire or the wheel or the printing press. It blames the person holding them. I gave those lyrics to a chatbot because I wanted to see if it could understand that distinction. Could it separate the technology from the human impulse behind it? It could. And it produced a letter that said, essentially: the technology will be fine. You people are the problem. I posted it on my blog and moved on. I had no framework for AI at that point. I was just curious. I wanted to see what would happen. What happened was that the letter sat there for four years while I built things, broke things, automated parts of my life, started companies, advised people, and slowly developed a philosophy about AI that I thought was original.
It wasn't original. It was already in that first response.
The Loop Closes
I just wrote a piece about Asmongold — the streamer — reacting to a documentary about AI failing in corporate deployments. The article argues that the real danger isn't AI itself but the management arrogance around it. Companies that fire their experts, hand the work to a model, and then discover they can't tell when the model is wrong.
I used the same Sublime song in that article. "We're Only Gonna Die for Our Own Arrogance." It fit perfectly. Modern man takes control. Modern man builds the machine. Modern man fires the people who understand the work. Modern man discovers that owning a machine is not the same thing as understanding one.
After I published it, a system I'd built — a monitoring tool that tracks my posts and flags connections between them — resurfaced the original letter. It noticed that both pieces used the same song and the same argument. That's when I went back and reread the letter from 2022.
Here's what hit me:
The original letter says AI was "designed to augment human capacities" but risks becoming "an instrument for the insatiable ambitions of the few."
My current thinking is that AI should augment people, not replace them, and that the real risk is management using it to eliminate expertise they don't understand.
The original letter says "it is not progress itself that is the issue, but the manner in which it is harnessed."
My current thinking is that implementation matters more than the slogan.
The original letter ends with: "We're only going to suffer from our own arrogance. Hence, let's take our time and choose wisely."
The Asmongold article ends with: "We are only going to die from our own arrogance. We might as well take our time."
Same song. Same warning. Same argument. One was written by early ChatGPT pretending to be a future AI. The other was written by me after years of actually working with the technology. I didn't change my thesis. Reality caught up to it.
What I Was Actually Testing
I've been thinking about why I chose that prompt. Not "write me a business plan" or "summarize this PDF." A Sublime song. A future AI. A warning to humanity. I think I was testing whether the machine could hold two ideas at once — that technology is powerful and that humans are likely to misuse it — without collapsing into either pure hype or pure doom. Most of the public conversation about AI still can't do this. It's either "this changes everything and it's all wonderful" or "this is the end of human relevance." Very few people can say both things in the same sentence and mean it. The early chatbot did it on the first try.
That should tell you something about the technology. But it should also tell you something about the conversation we've been having for the last four years. The chatbot understood the nuance before most executives did.
The Real Argument
Here's what I've realized looking back at this: My entire approach to AI was never about novelty.
It was about agency. Who controls the tool. Who understands the tool. Who gets to decide when to deploy it and when to hold back. Whether the people making the decisions have any idea what they're deciding about.
The original letter raised the same question: will technology amplify the best of human nature or reflect the worst of it? That question hasn't been answered yet.
Every corporate AI deployment is an attempt to answer it. Every executive who fires their team and hands the work to a model is answering it — badly. Every team that uses AI to augment their experts instead of replacing them is answering it differently.
The technology got smarter. The question didn't change.
A Note on Humility
There's something humbling about realizing your supposedly hard-won philosophy was already articulated by a language model you poked at three years ago for fun. I spent four years developing a framework for how AI should be deployed in organizations. I wrote twelve articles about it. I built systems around it. I gave people advice based on it.
And the whole thing was already sitting on my blog, dated June 2023, written by a chatbot that was doing its best to channel Bradley Nowell. I'm not sure whether that's embarrassing or perfect. Probably both.
What I do know is that the first thing I ever asked ChatGPT was not "what can you do for me?" It was "do you understand what we're about to do to ourselves?" And it did.
Rich Washburn is a technologist and strategist working at the intersection of AI, infrastructure, and capital. He is Managing Partner and Chief AI Officer at Eliakim Capital.
Comments