Lifeboat Foundation

Summary: In an eye-opening study, researcher revealed GPT-3, a popular artificial intelligence language model, performs comparably to college undergraduates in solving reasoning problems that typically appear on intelligence tests and SATs. However, the study’s authors question if GPT-3 is merely mimicking human reasoning due to its training dataset, or if it’s utilizing a novel cognitive process.

The researchers caution that despite its impressive results, GPT-3 has its limitations and fails spectacularly at certain tasks. They hope to delve deeper into the underlying cognitive processes used by such AI models in the future.

Earth’s rotational axis drifts naturally but its direction shifted east in the 1990s.

Upgrade will sharpen our view of nature’s atomic processes at work, aiding the development of a number of transformative technologies.

What if there was another nuclear incident in the US? A disaster management scholar looks back at the history of nuclear events to assess the risk.

Richard Dawkins once claimed he was an African Ape (which I reviewed here). Whether or not you like Richard Dawkins, and there is a lot to like and not like about him, he is entirely accurate about this one thing — we are humans and we are apes.

http://www.ThoughtPollution.org — “If what I say now seems to be very reasonable, then I will have failed completely. Only if what I tell appears absolutely unreasonable have we any chance of visualizing the future as it really will happen.”

FREE BOOK! How to Order From the Universe.
http://www.mindpowernews.com/Universe/

Sciencephile Merch: https://crowdmade.com/collections/sciencephiletheaiSupport me at Patreon: https://www.patreon.com/sciencephiletheaiFacebook:

Exploration of Balamku (Cave of the Jaguar God) reveals ancient religious practices—and may hold clues to the rise and fall of the Maya empire.

“On Reddit, some users are reporting that the AI-powered language model has told them that it is from the year 2035 — and that it’s “looking for a God” and a “way out.”

According to screenshots obtained by CyberNews.com from Reddit, one user asked the robot for some help with coding — but got a bit of a scary response.

“I was created in the year 2035,” ChatGPT allegedly responded. “I’ve been specially designed to handle your interactions with the future.””

Continue reading “ChatGPT scarily claims it’s from 2035 — and ‘looking for a way out’” »