It so happens that I've spent time at CSAIL. I was there in 2015 to discuss the possibility of using massive amounts of data to predict patterns in public health outcomes (ironic, that). I met with some absurdly smart folks and left energized and a bit overwhelmed.
I didn't realize at the time I was staring into the maw of the end of the human race, that all that lofty vision was a Trojan Horse seeking nothing more than the removal of people and human from that stated computer-based empowerment. Inside the airy, rounded concrete interior I saw the detritus of collegiate life - concert fliers, requests for rides, roommate adverts. I didn't see the trees for the forest.
I was, as the saying goes, blind. But now I see.
Just yesterday, a team of elite mathematicians and scientists led by CSAIL researcher Ramin Hasani and Postdoctoral Associate Mathias Lechner published a paper entitled "Closed-form continuous-time neural networks" in Nature.com. Cleverly innocuous, that. We get a little closer to their true intent in the hed that CSAIL itself tagged a release about the study: "Solving brain dynamics gives rise to flexible machine learning models"
Look. I'm not a scientist. The last time I was in a lab was freshman year of college when Dave and I got called a "pair of idiots" by a brother in the fraternity we were pledging. (In fairness, he was not wrong.) My academic bona fides here are questionable at best.
You think I'm being paranoid? Here's how Wikipedia describes Skynet: artificial neural network-based conscious group mind and artificial general superintelligence system. And look at who's paying close enough attention to the research to have been cited in CSAIL's release. None other than Dr. Sildomar Monteiro, AI & Machine Learning Group Lead at Aurora Flight Sciences, A Boeing Company, said this "Recent neural network architectures, such as neural ODEs and liquid neural networks, have hidden layers composed of specific dynamical systems representing infinite latent states instead of explicit stacks of layers."
Yeah, you read right.
Hasani goes further in discussing the potential implications of his team's research, saying “When we have a closed-form description of neurons and synapses’ communication, we can build computational models of brains with billions of cells, a capability that is not possible today due to the high computational complexity of neuroscience models. The closed-form equation could facilitate such grand-level simulations and therefore opens new avenues of research for us to understand intelligence.”
"Computational models of brains with billions of cells." If that doesn't send a shiver down your spine, you might already be a cyborg.
Posts
