The OpenAI/Scarlett Johansson scandal - Hidden aspect

➤ Background

👉 OpenAI aims to create Artificial General Intelligence (AGI). They stated this publicly many times.

👉OpenAI's most recent scandal:

OpenAI’s new model, GPT-4o, reads things out loud. One of the voices they created, “Sky”, is very similar to the voice of Scarlett Johansson, who previously declined their offer to narrate the system. OpenAI took it down after letters from her lawyers. Johansson issued a detailed and biting press release.

➤ But why want Johansson's voice?

🤔Hypothesis: Because she was the voice of an AGI in the movie Her.

🤔From Johansson's press release:

"Mr Altman even insinuated that the similarity was intentional, tweeting a single word 'her' [during the demo] - a reference to the film in which I voiced a chat system, Samantha, who forms an intimate relationship with a human."

🤔Altman's tweet is revealing. 

Even putting aside whether they tried to mimic Johansson's voice, they want their product to feel like a science fiction entity: Artificial General Intelligence as we know it from movies.

➤ Not the first time & Dangerous 

😵‍💫OpenAI’s core design choices also create AGI illusion (whether or not intended as such)

😵‍💫Did you ever wonder why ChatGPT spits out sentences in the first person?

😵‍💫Sentences in the first person create and reinforce the impression of communicating with a person, especially when they come without any clarification.

😵‍💫The illusion is powerful. We hear from time to time that even some software developers fall into this trap and attribute human-like agency to chatbots. 

😵‍💫 This is dangerous. It attributes too much power to nonsense LLMs spit out. It also distracts the masses from the real risks of LLMs and OpenAI’s responsibility. 

➤ Don't fall into the trap

⚠️Fortune tellers are great at making people feel like they know facts about their clients' past and future. It's a hoax.

⚠️ChatGPT (and similar chatbots) are great at making people feel like they are an intelligent entity.

⚠️It's a hoax: OpenAI is creating psychological cues for an AGI illusion.

➤ Join the conversation on the LinkedIn thread!