- One of the most relevant and often quoted topics in the present times is about the generative Alternative Intelligence that has stormed the collective consciousness of the global community unprecedentedly. The wide-ranging benefits that the latest cutting-edge technology offers humankind are unbelievable as well as unfathomable. The power of the information technology-driven measures has already helped humanity immensely. Nonetheless, the astonishing opportunities offered by AI are simply mind-boggling. Little wonder, every other Big Tech is in the fray to further explore the almost infinite possibilities offered by AI to even better humanity in more than one way. Grokipedia, from the stable of Musk, is one such new endeavour. Or is it?
PC: NBC News
- Musk has launched Grokipedia, saying that its goal is the truth, the whole truth, and nothing but the truth. Musk says that truth is the foundation of his Grokipedia. But how can anyone trust AI to know what’s true and what’s not? For the uninitiated, in Western thought, Aristotle’s correspondence theory is the gold standard definition of truth and falsity. Truth corresponds to a fact in the world; falsity doesn’t. That definition is 23 centuries old. But deciding what is true and what is false, or agreeing on this, has certainly not become easier since then. Mind you, humans have continued to set a lot of store by this endeavour. Modernity has been as much about the determination of measurable, verifiable, universal truths as individuality and subjectivity.
PC: Free Job Alert
- It is in the language of this hoary tradition that Elon Musk has launched Grokipedia. Except, darn it, the main qualification of this exercise seems to be the elimination of human touch. Grokipedia has been launched as the Other of Wikipedia. That it has wholesoul plagiarized Wikipedia is ironic, but also completely of a piece with how AI is degrading human copyrights. As we know, Wikipedia’s USP is transparent edit histories and messy human arguments, while Grokipedia’s is being all-AI. Of course, Wikipedia often stumbles in telling truths, but it is humans who bicker over it. Think about handing this complex matter to an LLM. These language models don’t even understand their system prompts precisely, let alone the concepts of truth and falsehood.
PC: Cybernews
- Note that, a couple of years ago, researchers estimated that AI chatbots can hallucinate anywhere from 3% to 27% of the time. So, to trust knowledge creation to AI’s statistical trick of predicting the next word is a mug’s game. The way everyone on X is hunted by Musk’s posts, unless they block him, betrays another dangerous kink in his encyclopaedia. The world’s richest man wants discourse dominance and X, Grok, and Grokipedia; they can all be rigged to fall in line. Therein lies the catch. Chatbots have shown themselves overeager to please even plain vanilla users. Imagine how much more sycophantic LLMs can be to their patrons. Don’t be fooled that this is just a culture war. The entire corpus of human knowledge is at stake here. Watch out for more info on this
