This week:
- Continuing discussion on Meta's Llama 3 open-source release - "Meta spent almost as much as the Manhattan Project on GPUs in today’s dollars."
- Microsoft releases their Work Trend Index claiming 75% of "global knowledge workers" are using AI. I think the most striking section of this report is regarding benefits to power users.
- Adobe believes stock photography will quickly become obsolete (ok!), and Photo AI 3 is another tool for photographers to "denoise" and enhance grainy photography.
- More open-source tools powered by AI for mass web scraping, along with previous releases like Google's ScreenAI, are (to me) one of the most exciting new frontiers being developed and will represent a shift in how we engage with the world wide web!
- OpenAI addresses their "approach to shaping desired model behavior and how we evaluate tradeoffs when conflicts arise."
- Large language models versus 180 students of counseling psychology on social intelligence.
- "Introducing AlphaFold 3, a new AI model developed by Google DeepMind and Isomorphic Labs. By accurately predicting the structure of proteins, DNA, RNA, […] we hope it will transform our understanding of the biological world and drug discovery."
- And for something I am hoping to learn more about this week - a new forecasting model from Google: "TimesFM is a forecasting model, pre-trained on a large time-series corpus of 100 billion real world time-points, that displays impressive zero-shot performance on a variety of public benchmarks from different domains and granularities."
- And lastly - a sentiment I fully agree with: "But I’m increasingly of the opinion that sharing unreviewed content that has been artificially generated with other people is rude." I think Simon Willson's personal AI ethics are timely and mostly spot-on (for now). Putting pen to paper here is highly recommended (and something I plan to share).