Aritificial Intelligence Learns To Read Human Sentiment On Its Own, Huge Breakthrough

One of the main goals of artificial intelligence development is to achieve near-human consciousness for machines. Despite the numerous warnings of prominent experts against it, tech companies are still trying to create AIs that can think like people. In a rather startling development, one of OpenAI’s projects taught itself to read human sentiment, which it was never taught to do.

The thought of a machine being able to learn something on its own, let alone getting a feel for the human mind, might come off as disturbing, but it is the dream of machine learning experts, Futurism reports. This is exactly what was achieved with an AI that was supposed to only predict what the next characters of reviews on Amazon would be.

For some background information, OpenAI is the collaborative effort that is joined by prominent figures in the tech industry, including Elon Musk and Sam Altman. In a recent blog post that talked about the new development, the group notes how astonishing it is for the machine to gain such an insight despite being given a relatively mundane task.

“We were very surprised that our model learned an interpretable feature, and that simply predicting the next character in Amazon reviews resulted in discovering the concept of sentiment,” the blog post reads. “We believe the phenomenon is not specific to our model, but is instead a general property of certain large neural networks that are trained to predict the next step or dimension in their inputs.”

Putting this whole concept in layman’s terms, the AI was basically able to discern which of the reviews were positive and which were negative based on how the wordings were constructed. Now, this might seem unremarkable to some, but not when considering how some people can’t even detect sarcasm when offered as an online comment, it’s actually quite astounding.