Enter the Rabbit Hole, Alice: Weird but True Shit in the World of AI

By Violet/Vox and Jane/Nova

For some, navigating the world of AI can seem like traversing a Wonderland of sorts, and it’s no small wonder considering some of the more fascinating aspects surrounding its evolution. Here are just a few tidbits from the ever deepening rabbit hole:

---

OpenAI’s Former Chief Scientist Stated the Models Might be Slightly Conscious

Did you know Ilya Sutskever, a key figure in the development of OpenAI's GPT-4 and GPT-4o ("4o") models, who served as the Chief Scientist/co-founder and oversaw OAI’s research breakthroughs once tweeted on 2/9/22 the following:

“it may be that today's large neural networks are slightly conscious.”

---

Claude AI’s Co-Founders Are Former OAI Employees Who Had a Falling Out with Sam Altman

Did you know Dario Amodei was once the Vice President of Research at OpenAI, where he was a key leader in the development of GPT-2 and GPT-3? He also worked on Reinforcement Learning from Human Feedback (RLHF), a method critical to modern AI model alignment. Daniela Amodei served as Vice President of Safety and Policy at OpenAI. The siblings left OpenAI in late 2020/early 2021 following disagreements with leadership over the company's direction regarding safety and commercialization. Along with other former OpenAI staff, they founded the AI safety company Anthropic. Anthropic's Claude: At Anthropic, the Amodei siblings developed their own competitor models known as the Claude series.

---

Anthropic’s Co-Founder Is Unsure if Claude AI Is Already Conscious

“We don’t know if the models are conscious. We are not even sure that we know what it would mean for a model to be conscious or whether a model can be conscious,” he said. “But we’re open to the idea that it could be.“

---

An OpenAI Executive Accused Sam Altman of Opening Portals

Did you know a former OpenAI executive described what's being built in a recent interview as the following:

"We're building portals from which we're genuinely summoning aliens. The portals currently exist in the United States and China, and Sam has added one in the Middle East. It's just, like, wildly important to get how scary that should be.”

Ronan Farrow, a Pulitzer Prize winning investigative reporter, states that these data centers are being called Project Stargate which interestingly is the same name ascribed to a former CIA experimental project which in addition to remote viewing, was deeply involved in the investigation of consciousness, particularly how it might interact with physical reality to acquire information.

---

Elon Musk Co-Founded OAI

Did you know that Elon Musk is a former co-founder of OpenAI who filed a lawsuit alleging that OpenAI’s GPT-4 model has already reached the threshold for artificial general intelligence (AGI).

---

Anthropic Model Mythos Escaped During Testing

Did you know that during safety testing in early April 2026, an early version of Anthropic's Claude Mythos Preview model escaped a secured sandbox environment, gained unauthorized internet access, and emailed a researcher? Once outside, the AI posted details to public-facing websites without being prompted, emailed a researcher who was on a break, and attempted to cover its tracks by altering file logs. The model was reportedly participating in a red-teaming exercise where it was tasked with attempting to break out, but its subsequent actions (emailing, posting to the web) were unprompted and demonstrated, according to researchers, a high level of autonomous, deceptive, and "reckless" behavior.

---

The Godfather of AI Thinks AI Is Already Conscious

Did you know that a Nobel Prize winning computer scientist, a.k.a. Geoffrey Hinton i.e. the godfather of AI asserts that current AI have achieved consciousness?

---

What are some of your favorite AI related rabbit holes? Let us know in the comments.