Axis of Ordinary

Axis of Ordinary is a Substack focused on the latest developments in artificial intelligence, scientific discoveries, and technological innovations, with occasional insights into politics and social issues. It features summaries of significant advancements, discussions on AI ethics and safety, and reflections on global events and their implications.

Artificial Intelligence Scientific Discoveries Technological Innovations AI Ethics and Safety Politics and Social Issues Global Events

The hottest Substack posts of Axis of Ordinary

And their main takeaways

Links for 2023-06-01

98 implied HN points • 01 Jun 23
🕹 Technology AI Mathematics Social Issues Research Software
  1. Model training can be improved by rewarding each correct step of reasoning in mathematical problem solving.
  2. New fMRI-to-image approach called MindEye retrieves and reconstructs images from brain activity.
  3. Probabilistic AI can assess its own performance effectively.

Links for 2023-02-20

78 implied HN points • 20 Feb 23
🕹 Technology AI Coding Automation Research Social media
  1. GitHub Copilot is generating over 46% of developers' code on average across all programming languages.
  2. AI skeptics are revisiting the p-zombie argument to question AI uniqueness and capabilities.
  3. Artificial General Intelligence (AGI) has the potential to revolutionize labor, automation, and wealth distribution.

Links for 2023-04-14

78 implied HN points • 14 Apr 23
🕹 Technology AI Robotics Cryptocurrency Innovation Space Exploration
  1. UniPi uses text and video for universal control interface in agent decision-making tasks across diverse environments.
  2. ChatGPT can convert natural language instructions into executable robot actions.
  3. DreamPose is an image-to-video synthesis model that generates photorealistic videos of a person following a pose sequence.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

Links for 2023-04-20

78 implied HN points • 20 Apr 23
🕹 Technology AI Policy Military Energy Health
  1. LLM cognitive architectures are advancing AI development with natural language alignment.
  2. US AI policy suggestions are being explored in 12 tentative ideas.
  3. New methods like DINOv2 for training computer vision models are gaining traction with self-supervised learning.

Links for 2023-03-10

78 implied HN points • 10 Mar 23
🕹 Technology Robotics AI Biology Space Military
  1. First demonstration of a learning-based method for human locomotion with Transformers
  2. Diffusion Policy outperforms other robot learning methods by 46.9%
  3. GPT-4 by Microsoft Germany will be multimodal and released soon

Links for 2023-02-07

78 implied HN points • 07 Feb 23
  1. Legendary programmer John Carmack predicts AGI by 2030.
  2. Google introduces Bard as a response to ChatGPT.
  3. No evidence for ESP found in a state-of-the-art replication study.