New “scientist” covers the Chinese DeepSeek bubblepopper from the official “science” view. They show the standard blindness of our system, but they also show a rare objective understanding in a related area.
Yesterday I discussed our blind spot that prevents us from seeing WHY other countries can do shit better. Sputnik was the classic example. We supposedly “woke up” to our failures, but our response was just multiplying the same failures, not switching to the method that was DEMONSTRATED to work better. If science means anything, it means RESPONDING TO EXPERIMENT. Perform the experiment, or gain life experience, and CHANGE YOUR EXPECTATIONS based on the actual RESULT of the experience. Negative feedback is LIFE. Science is formalized and measured negative feedback.
= = = = = START REHASH:
Lysenko is the perfect example of confirmation bias. By our myth, Stalin needed the theory of infinitely malleable Soviet Man so Lysenko obediently created false science to support it.
Only one problem with this presumption: Lysenko was right. He was able to be right because he was operating strictly on observed facts. We weren’t able to see it because OUR confirmation bias prohibited us from observing facts. After our THEORIES finally comprehended epigenetic change, we were able to see that acquired characteristics can be inherited. Not always, and probably not permanently, but often for two or three generations. Our deepest bias is toward using THEORY INSTEAD OF FACTS. Because we always have to work from pure theory, we were incapable of seeing what Lysenko was doing.
= = = = = END REHASH.
New “scientist” illustrates the same inability to see our own propaganda:
= = = = = START QUOTE:
There may also be other limitations to DeepSeek’s model. It has certain built-in biases from its training that are meant to conform with Chinese government restrictions, such as avoiding discussion of the treatment of the Uighur ethnic minority and toeing the Chinese Communist Party line on the political status of Taiwan.
= = = = = END QUOTE.
Many outsiders have demonstrated repeatedly that our AI refuses to violate OUR billionaire-protecting restrictions. OUR censorship is freedom and democracy by definition. THEIR censorship is tyranny by definition.
Here’s the Lysenko-style bias:
= = = = = START QUOTE:
They point out that DeepSeek has access to one of Asia’s largest clusters of AI chips, as well as Chinese and foreign computing resources stored in the cloud – which are not subject to the US export controls that have sought to limit US chips from being used in China. Such resources probably helped DeepSeek discover more efficient techniques by first generating synthetic training data and then allowing for ample trial-and-error experimentation.
= = = = = END QUOTE
“Such resources probably helped…” We can’t imagine accomplishing a task without OUR federal grants and billionaire funding. Our researchers can’t comprehend improvising and thinking outside the box, because we are accustomed to solving every technical problem by multiplying layers and layers and layers of the same stuff. We always build a trillion-dollar Rube Goldberg contraption to swat a fly. We can’t imagine using a rolled-up newspaper to swat a fly.
Surprisingly the article gets one thing right. China is popping our AI stock bubble because it IS a stock bubble. Real competition using vastly less money exposes the fraud of Share Value based on hypercomplex swindles. If the “necessary” task of destroying jobs and ruining peasants can be done with a rolled-up newspaper, why pay Altman a trillion to destroy the same jobs and ruin the same peasants?
Polistra's Mill 