Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

Is Aschenbrenner's 165 page paper on AI the naivety of a 25 year old?

Jun 13, 2024 - fikisipi.substack.com
The author discusses their initial skepticism towards neural networks in data science, believing that the human brain's complexity could never be approximated by a computer. They cite reasons such as the brain's hundreds of millions of years of development, complex synapses, difficulty in scanning, and the intricate role of neurotransmitters. However, their perspective changed in 2019 when they interacted with AI Dungeon, a text game based on GPT-2, which provided an uncannily accurate response to their input.

The author then shares a link to a paper titled "Situational Awareness" by Leopold Aschenbrenner, an ex-OpenAI employee, who discusses the potential successes and problems that AI progress may cause. The author concludes by sharing two random facts: the progress in solving Riemann’s Hypothesis, a 165-year-old unsolved math puzzle, and the creation of a mail stamp celebrating the birth of Claude Shannon, a key figure in the foundation of artificial intelligence.

Key takeaways:

  • The author initially had a negative view of neural networks in data science, believing the human brain's complexity could never be approximated by a computer.
  • The author's perspective changed after interacting with AI Dungeon, a text game based on GPT-2, which provided an uncanny and immersive experience.
  • The author recommends a paper named 'Situational Awareness' by Leopold Aschenbrenner, an ex OpenAI employee, which discusses the potential successes and problems AI progress may cause in the future.
  • The author shares two random facts: the progress on Riemann’s Hypothesis by two mathematicians and a mail stamp celebrating the birth of Claude Shannon, a key figure in the foundation of artificial intelligence, proposed by a designer from the author's country of origin.
View Full Article

Comments (0)

Be the first to comment!