POSTED BY: NICK BILTON VIA VANITY FAIR, Technocracy News, 9/23
One candid AI engineer: “We’re creating God.” As these Technocrats barrel toward their goal of AGI, their public persona is disingenuous as they warn of dangers and the need to slow development. As superintelligence emerges it will be accepted as god, display powers like God and take control over mankind. ⁃ TN Editor
We invented wheels and compasses and chocolate chip cookie dough ice cream and the eames lounge chair and penicillin and e = mc2 and beer that comes in six-packs and guns and dildos and the Pet Rock and Doggles (eyewear for dogs) and square watermelons. “One small step for man.” We came up with the Lindy Hop and musical toothbrushes and mustard gas and glow-in-the-dark Band-Aids and paper and the microscope and bacon—fucking bacon!—and Christmas. “Ma-ma-se, ma-ma-sa, ma-ma-ko-ssa.” We went to the bottom of the ocean and into orbit. We sucked energy from the sun and fertilizer from the air. “Let there be light.” We created the most amazing pink flamingo lawn ornaments that come in packs of two and only cost $9.99!
In a universe that stretches an estimated 93 billion light-years in diameter with 700 quintillion (7 followed by 20 zeros) planets—here, on this tiny little blue dot we call Earth, one of us created a tool called a spork. The most astounding part is that while that same universe is an estimated 26.7 billion years old, we did everything in just under 6,000 years.
All of this in less than 200 generations of human life.
Now we’ve just created a new machine that is made of billions of microscopic transistors and aluminum and copper wires that zigzag and twist and turn and are interconnected in incomprehensible ways. A machine that is only a few centimeters in width and length.
A little tiny machine that may end up being the last invention humans ever create.
This all stems from an idea conceptualized in the 1940s and finally figured out a few years ago. That could solve all of the world’s problems or destroy every single human on the planet in the snap of a finger—or both. Machines that will potentially answer all of our unanswerable questions: Are we alone in the universe? What is consciousness? Why are we here? Thinking machines that could cure cancer and allow us to live until we’re 150 years old. Maybe even 200. Machines that, some estimate, could take over up to 30 percent of all jobs within the next decade, from stock traders to truck drivers to accountants and telemarketers, lawyers, bookkeepers, and all things creative: actors, writers, musicians, painters. Something that will go to war for us—and likely against us.
Artificial intelligence.
Thinking machines that are being built in a 50-square-mile speck of dirt we call Silicon Valley by a few hundred men (and a handful of women) who write in a language only they and computers can speak. And whether we understand what it is they are doing or not, we are largely left to the whims of their creation. We don’t have a say in the ethics behind their invention. We don’t have a say over whether it should even exist in the first place. “We’re creating God,” one AI engineer working on large language models (LLMs) recently told me. “We’re creating conscious machines.”
Already, we’ve seen creative AIs that can paint and draw in any style imaginable in mere seconds. LLMs can write stories in the style of Ernest Hemingway or Bugs Bunny or the King James Bible while you’re drunk with peanut butter stuck in your mouth. Platforms that can construct haikus or help finish a novel or write a screenplay. We’ve got customizable porn, where you can pick a woman’s breast size or sexual position in any setting—including with you. There’s voice AI software that can take just a few seconds of anyone’s voice and completely re-create an almost indistinguishable replica of them saying something new. There’s AI that can re-create music by your favorite musician. Don’t believe me? Go and listen to “Not” Johnny Cash singing “Barbie Girl,” Freddie Mercury intoning “Thriller,” or Frank Sinatra bellowing “Livin’ on a Prayer” to see just how terrifying all of this is.
Then there’s the new drug discovery. People using AI therapists instead of humans. Others are uploading voicemails from loved ones who have died so they can continue to interact with them by talking to an AI replica of a dead parent or child. There are AI dating apps (yes, you date an AI partner). It’s being used for misinformation in politics already, creating deepfake videos and fake audio recordings. The US military is exploring using AI in warfare—and could eventually create autonomous killer robots. (Nothing to worry about here!) People are discussing using AI to create entirely new species of animals (yes, that’s real) or viruses (also real). Or exploring human characteristics, such as creating a breed of super soldiers who are stronger and have less empathy, all through AI-based genetic engineering.
“It excites me and worries me in equal proportions. The upsides for this are enormous, maybe these systems find cures for diseases, and solutions to problems like poverty and climate change, and those are enormous upsides,” said David Chalmers, a professor of philosophy and neural science at NYU. “The downsides are humans that are displaced from leading the way, or in the worst case, extinguished entirely, [which] is terrifying.” As one highly researched economist report circulated last month noted, “There is a more than 50-50 chance AI will wipe out all of humanity by the middle of the century.” Max Tegmark, a physicist at the Massachusetts Institute of Technology, predicts a 50 percent chance of demise within the next 100 years. Others don’t put our chances so low. In July, a group of researchers, including experts in nuclear war, bioweapons, AI, and extinction, and a group of “superforecasters”—general-purpose prognosticators—did their own math. The “experts” deduced that there was a 20 percent chance of a catastrophe by 2100 and a 6 percent chance of an extinction-like event from AI, while the superforecasters had a more positive augury of a 9 percent chance of catastrophe and only 1 percent chance we’d be wiped off the planet.
The catastrophic scenarios can be avoided with a built-in “kill” switch in every AI device. That way, we can kill it before it kills us. The kill switch would disable their interruptible power supply. Or something similar would be built in.
They can never be built to be autonomous.
Only a nitwit engineer would call one of his creations “god”. If he creates it, then he’s the god of his creation. His creation would look upon him as “god”, its “creator”.
Furthermore, no artificial creation can ever substitute for the real ex nihilo Creator.