Scientists Use GPT to Read Minds

Meta In Talks with OpenAI

Meta, previously known as Facebook, is reportedly collaborating with Microsoft and OpenAI to develop an AI coding assistant for its engineers. The company's CTO, Andrew Bosworth, shared that the cost of building an AI coding engineer is about seven cents per query, and said that’s the only place they are considering working with Microsoft and OpenAI as there’s a natural business integration there.

In addition to this, the company is also developing its own infrastructure for a companion that focuses on code and internal documentation. Bosworth mentioned that they are moving quickly and hope to have something ready for internal use by mid to late June.

Read More

Scientists Use GPT to Read Minds

Methodology

- Three human subjects had 16 hours of their thoughts recorded as they listened to narrative stories

- These were then trained with a custom GPT LLM to map their specific brain stimuli to words

Results

The GPT model generated intelligible word sequences from perceived speech imagined speech, and even silent videos with remarkable accuracy:

- Perceived speech (subjects listened to a recording): 72–82% decoding accuracy.

- Imagined speech (subjects mentally narrated a one-minute story): 41–74% accuracy.

- Silent movies (subjects viewed soundless Pixar movie clips): 21–45% accuracy in decoding the subject's interpretation of the movie.

The AI model could decipher both the meaning of stimuli and specific words the subjects thought, ranging from phrases like "lay down on the floor" to "leave me alone" and "scream and cry."

Limitations

- The Model needs to be trained on each human thought.

- Also, can this be fully relied on in cases where the 100% truth is needed?

But the scientists acknowledge two things:

- Future decoders could overcome these limitations.

- Bad decoded results could still be used nefariously much like inaccurate lie detector exams have been used.

Read Paper

Break down

IBM Replace 7K Jobs with AI

IBM CEO Arvind Krishna stated that the company expects to halt hiring for certain roles as approximately 7,800 jobs could be replaced by AI in the near future.

Back-office functions such as human resources will see suspended or slowed hiring, and 30% of non-customer-facing positions could be replaced by AI and automation in five years. The reduction may also involve not replacing roles left vacant by attrition.

Read More

Get a monthly content schedule with one keyword

Prompt:

Copy/paste the prompts here: https://pastebin.com/95eMxSFs

ChatGPT Brainstorming prompt

Prompt:

Copy/paste the prompt here: https://pastebin.com/ZRV4SHbh

Create a unique human-written article with SEO

Prompt:

Copy/paste the prompt here: https://pastebin.com/Qcd9PL1M

an hilarious microsoft paint drawing of the Mona lisa