@0xbitcoiner kicked off an interesting sidequest this week with: OpenAI Ramps Up Robotics Work in Race Toward AGI, as somehow OpenAI has silently let go of their previous method of reaching AGI through massive datacenters under the banner of bigger is better, and now claims that robots are needed for AGI. The Fever Dream of Imminent Superintelligence Is Finally Breaking was shared by @Scoresby which underlines that governments, markets and plebs that paid to have their most private thoughts stolen have been falsely lead into this belief that AGI is imminent.

In case you thought that OpenAI does not read your stuff, @k00b shared graph porn about How People Use ChatGPT and @Cje95 shared the underlying article: OpenAI Releases First-of-Kind Study On How Use Of ChatGPT For Everyday Tasks. Yes, they know everything you tell the bot. But they didn't publish stats about how much of the information provided was factually correct. Plausibility is not truth, @Scoresby shared, as well as an explainer about Magical Thinking on AI.

Some still believe in AGI: Which Human Tasks Will Become More Popular When Computers Become Unbeatable?, @kr asks, and Ilya Sutskever stubbornly believes his own gaslighting: AI will replace all human labor, shared by @lunin

Why OpenAI’s solution to AI hallucinations would kill ChatGPT tomorrow was shared by @south_korea_ln, which would be a great outcome, as @SimpleStacker reports that ChatGPT tried to gaslight me about OpenAI reasoning token billing.

But it's not just OpenAI: Millions turn to AI chatbots for spiritual guidance and confession was shared by @0xbitcoiner even though the Pope says nope to AI pope. Artificial Spirituality sounds like a contradiction.

It was a terrible week for AI-related privacy!