Of course I believe in cause and effect; I’m just uncertain which comes first
Watching Jensen Huang’s keynote at CES this week made me hit the rewind button nearly a dozen times: did he say petaflops?? (starting 2:30) Yes, it turns out, petaflops. Four of them. How is this even humanly possible? After all, let’s remind ourselves, a petaflop is one quadrillion floating-point operations (FLOPs) per second.* Sure, I know what a second is, roughly, but what the hell is a quadrillion? Also, the same guy just moments before was touting his fabled taste in leather jackets—as if his jackets were also only possible because of AI, signaling his authority over the massive audience. And then you even had Gary Shapiro introducing Mr. Huang by mispronouncing “Nvidia” repeatedly, even though he claims Nvidia is “one of the most consequential companies in the world.”
Hmm. You’re telling me that humans—even these two humans—work on a peta- (1,000,000,000,000,000) scale with ease? It sounded so surreal, I started to think those devices in Mr. Huang’s hands weren’t flop’ing at all—they were some kind of communication devices to another dimension.
Then of course came the news that you can force one cell into another (called endosymbiosis) to create complex life. From the article:
Now, for the first time, researchers have watched the opening choreography of this microscopic dance by inducing endosymbiosis in the lab(opens a new tab). After injecting bacteria into a fungus — a process that required creative problem-solving (and a bicycle pump) — the researchers managed to spark cooperation without killing the bacteria or the host….The cells even adjusted to each other faster than anticipated. ‘To me, this means that organisms want to actually live together, and symbiosis is the norm,’ said Vasilis Kokkoris.
Pretty cool. This is certain to get the Nobel committee’s attention. But put yourself in the fungus’ perspective: what must it seem like to have an alien force artificially inject a biological entity into your body? And at a “Goldilocks rate” that ensures you and the bacteria “become addicted to each other”?
Well, now we know. Because isn’t that the description of the period we’re going through now? Are humans inventing artificial intelligence or is AI being introduced into our society by an invisible foreign actor? Maybe the real training that LLMs require is learning to think in our arrow of time? I find causality in these scenarios rather ambiguous.
Humanity has developed various, often conflicting versions of time. Our modern modality prefers the idea that we need to follow natural phenomenon as dominoes falling in such a way as to make sense to the human mind and, per Ana Bazac, “in the process of knowledge one can detect not only a beginning and an end, but also their terrestrial reasons, the changes, the conditions, the rhythm of changes and their new and new transformations.”* The author continues, “consequently, one can detect the direction of these transformations, and this direction … is that of the qualitative and quantitative acquisitions: of progress.”
What we’re witnessing may not be human-led progress at all, but an intervention brought on from who knows where. Will we become “addicted” to this technology or killed by it? I guess the lab coats in the sky are willing to find out. Here on earth, we’d better learn how to incorporate this artificially introduced intelligence in a symbiotic manner. I created AI for Society Online to explore how to traverse this epistemological break intact. The team has been hard at work at we plan to announce some of our work soon, so subscribe and stay in touch!