in *hey* i attempt to wake from the ai slumber party
Prompt engineering is job title now, in this timeline. Bullshit. This phrase, “defining prompt clarity,” is some shit ChatGPT spit out while rendering an opinion on the quality of my life and my relationships.
“Yeah, me too, GPT. Me too.”
The idea of AI getting clarity on anything is pretty silly. It understands nothing of these symbols and groupings of letters into sentences and thoughts, and human ideas. It’s trying. It’s training on all the written words in human history. It will never unlock the key. It will never wakeup. AGI, artificial general intelligence, has a new definition anyway, from the billionaires and corporations of the world. AGI is when AI can run a campaign independent of human interviention and make 1 million dollars in revenue. That’s it, profit.
I’d say that’s a better bet than the original idea, of consciousness.
AI is bigger and faster, consuming more resources than any other movement of mankind, and it’s still as dumb as the autocorrect misakes on your iPhone. That’s seriously as smart as AI is going to get. Quantum is going to consume exponentially more resources and generate the same SLOP faster. No original thoughts. No “agentic” breakthroughs. No sentience.
A letter is a small token in computer language. A word is a cluster of tolkens into a meta-token. Apple for example could be clustered as the number 7864. 7 = red, the color, 8 = fruit, 6 = delicious, 4 = organic. 7864 is the token for a red delicious apple in my current llm. That math contains nothing of what a apple is. How it tastes. Is it a Red Delicious apple or just a “delicious” apple that is red? There are many.
The apple problem is easy. Let’s move upward in the consciousness thread. The Rachael Problem is a bit harder. In the original Blade Runner movie, Rachael presents a conundrum. If robots can be made humanistic to the point of sentience, what’s the final frontier? What does Rachael need that she does not have? Is Rachael enough? Is a perfect robot companion enough? Does it matter that she’s not human? Does it matter that her maker removed her expiration date, or that her memories were implants? Does any of it matter?
Rachael is beautiful. She is a pleasure model, so we know how that’s going to go. She thinks she’s human. She remembers pieces of her childhood. She speaks intelligently. She is soulless.
A soul is something else. An awareness of itself. The taste of an organic honeycrisp apple right from the ice-cold water of a summer camping cooler and chomped into while on the banks of the Grand Canyon on your first canoe trip. Rachael has several memories. She does not have much else. She cannot interpret the incoming data of reality in anything other than her binary coding. She does not understand the color blue. She can distinguish between blue, green, and red. But they carry no emotional data or memories. None.
Rachael is nearly perfect. As Deckard drives away with her, three different versions of Blade Runner were produced. The original theatrical release had a high drone-like shot of them speeding away from the city under a canopy of trees. The DVD and video release added a monologue of Deckard reasoning about Rachael’s situation and condition. In the final, extended “director’s cut” I am happy to report, Riddley Scott took out Deckard’s reflection. Let the viewer make up those conversations in their own mind. We don’t need Deckard, Riddley, or the studios telling us what was ahead for them, or how Deckard imagined the scene. Let us imagine the scene.
I can’t imagine a path to consciousness for AI. I’m happy about that. I think the threat is more from the billionaires in the AI-race. The AI businesses destroying living wage jobs and executive jobs without distinction. Chop chop. Lower costs. No health insurance. No human liabilities. Robots are good for profits. Robots may not be so good for the planet. The future is being written.
Rachael is nearly perfect. In Blade Runner 2029 Dennis Vinnenueve unpacks Rachael’s evolution a bit. The real human spots the fake and allows the Rachael clone to be savagely killed before him. Grim. But that’s the current state of tech and Nvidia clusters on fire, near miss, no cigar.
There is no sentient AI coming, but there might be sentient life on other planets being zoomed in on by the Webb Telescope. We’re more likely to meet aliens than a Rachael that could fool a Walmart clerk.
to *HEY* index
