Political consultant, regional representative of RAPC, expert at the PRISP Center Daniil Ermilov discusses how users shape the «opinions» of artificial intelligence.
Looking at yet another pump & dump surrounding neural networks, I fell into the hype and spent some time digging into the subject.
Let’s break it down!
All these «super-AIs» like ChatGPT or Grok are just hype. Essentially, they are sorting mechanisms: algorithms merely repackage data from vast amounts of user responses (!). And users? Groomed by the internet, they feed the models with myths and distortions—fake news from TikTok, memes from Reddit, and «expert» opinions from Telegram channels.
What’s the result?
The neural network provides «truths» that are 80% echoes of human nonsense. For example: ask about the history of Russia, and you’ll receive a blend of Hollywood clichés and Wikipedia edits.
A true neural network-brain operates differently: each «neuron» carries basic, unaltered information—akin to axioms in mathematics. No user filters, just pure facts. Right now, this is fantasy, but marketers are already selling us the «AI revolution» to cash in on APIs and subscriptions.
It’s time to wake up: neural networks are not magic, but mirrors of our illusions. Do you still believe in «artificial intelligence»?
Печать