Links for 2022-01-30
Quantum Mechanics Must Be Complex: Two independent studies demonstrate that a formulation of quantum mechanics involving complex rather than real numbers is necessary to reproduce experimental results. https://physics.aps.org/articles/v15/7
“Last night I finally read "Infinite Hex is a Draw" -- one of the most elegant papers I've read in years. There's basically zero special notation, zero prerequisites, you can just read it straight through and stare at the pictures, and become convinced of one non-obvious conclusion after another after another about Hex on an infinite board. Plus there are many open problems that I'm risk of being nerd-sniped into thinking about!” (via a computer scientist) https://arxiv.org/abs/2201.06475
Papers unpicked: Strategy on an Infinite Chessboard between an Angel and a Devil https://www.youtube.com/watch?v=JvhSRCfCHb4
Let's remove Quaternions from every 3D Engine https://marctenbosch.com/quaternions/
Discussion Contests: “To induce informed discussion of a topic, BOTH subsidize a betting market on it, AND pay winners of an auction to talk on a center stage in proportion to the info added to market prices while they speak.” https://www.overcomingbias.com/2022/01/discussion-contests.html
No one really knows what happens inside an atom. [published January 02, 2020] https://www.livescience.com/mystery-of-proton-neutron-behavior-in-nucleus.html
The Dunning-Kruger Effect Is Probably Not Real https://www.mcgill.ca/oss/article/critical-thinking/dunning-kruger-effect-probably-not-real
Sars-CoV-2 spike protein activates human endogenous retroviruses in blood cells https://phys.org/news/2022-01-sars-cov-spike-protein-human-endogenous.html
Israeli Physicists Create Thought-provoking Model for Material That Never Melts https://www.haaretz.com/israel-news/MAGAZINE-israeli-physicists-create-thought-provoking-model-for-material-that-never-melts-1.10536581
"the 1127-30 Han migration southward in historic China had a significantly positive effect on economic prosperity in the year 2000 (as measured by GDP & nighttime lights per capita)" https://onlinelibrary.wiley.com/doi/abs/10.1111/iere.12565
Happiness increases with income and the relationship doesn't plateau at $75,000 https://www.pnas.org/content/118/4/e2016976118
The revolutionary drop in cost to get things (and sometimes even people) into space. A log scale chart. https://www.visualcapitalist.com/the-cost-of-space-flight/
Correction for yesterday’s link collection: #2 should have linked here https://bounded-regret.ghost.io/thought-experiments-provide-a-third-anchor/
I guess I should have noticed this earlier but unzipping in Python is equivalent to taking the transpose of a matrix:
inputMatrix = [ [7, 14, 21], [1, 2, 3] ]
transposeMatrix = zip(*inputMatrix)
print(list(transposeMatrix))
Output: [(7, 1), (14, 2), (21, 3)]
In the past, I have been very skeptical of the risks associated with artificial intelligence. But all my predictions have been shattered.
You can dismiss the answers shown in the screenshot above. You can make fun of it. But did you predict that this would be possible in 2022? Just 10 years ago, would you have expected this? I certainly didn't. And I don't feel confident to rule out further transformative breakthroughs in the very near future.
If you're worried about how climate change will affect us 50 to 80 years from now, imagine what half a century of progress in machine learning might mean.
No other issue comes close to the impact of artificial intelligence. This will transform everything forever and beyond recognition.
P.S. How confident are you that this sort of babbling is qualitatively different from human intelligence rather than quantitatively different; i.e. that it isn't just a matter of scaling and some fine-tuning? Confident enough to bet the world on it?
Would you have predicted that writing coherent text doesn't require intelligence and understanding? How confident are you that e.g. coming up with the plan for a doomsday weapon isn't similar; i.e. that it also doesn't require intelligence and understanding?