Links for 2023-03-05
“Formal verification is incredibly difficult. Writing proofs that code is correct often takes much longer than writing the code. PRoofster generates those proofs automatically!” https://people.cs.umass.edu/~brun/pubs/pubs/Agrawal23icse-demo.pdf (video: https://www.youtube.com/watch?v=xQAi66lRfwI)
“Two early papers find the effects of generative AI on knowledge work are completely unprecedented in modern history. Separate studies of both writers and programmers find 50% increases in productivity with AI, and higher performance and satisfaction. And this is just the start.” https://threadreaderapp.com/thread/1631397931604488194.html
High-resolution image reconstruction with latent diffusion models from human brain activity https://sites.google.com/view/stablediffusion-with-brain/
AI-Boosted Brave Browser Now Summarizes Search Engine Answers https://www.cnet.com/tech/computing/ai-boosted-brave-browser-now-summarizes-search-engine-answers/
Yudkowsky on Bankless follow-up Q&A: “Caring a very tiny fraction is basically the same problem as 100% caring. It's like, well, could you have a computer system that is usually like the Disk Operating System, but a tiny fraction of the time it's Windows 11? And writing that is just as difficult as writing Windows 11....you've got to think about AI technology and computers and humans and intelligent adversaries and distant superintelligences who might be trying to exploit your AI's imagination of those distant superintelligences and others like ridiculous weird problems that would take so long to explain.” https://www.lesswrong.com/posts/jfYnq8pKLpKLwaRGN/transcript-yudkowsky-on-bankless-follow-up-q-and-a
ChatGPT: A Meta-Analysis after 2.5 Months https://arxiv.org/abs/2302.13795
Flexible bioelectronics capable of tracking neural activity from the same cells during the entire adult life of mice. https://www.nature.com/articles/s41593-023-01267-x
Social Media is a Major Cause of the Mental Illness Epidemic in Teen Girls. Here’s the Evidence. https://jonathanhaidt.substack.com/p/social-media-mental-illness-epidemic
“Explaining the existence of such massive galaxies close to the dawn of time would require scientists to revisit either some basic rules of cosmology or the understanding of how the first galaxies were seeded from small clouds of stars and dust.” https://www.theguardian.com/science/2023/feb/22/universe-breakers-james-webb-telescope-detects-six-ancient-galaxies
South Korea Again Smashes Own Record for World’s Lowest Fertility Rate in 2022 [Bloomberg] https://archive.is/KPqRe
The South Korean government in the 1950s-60s embarked on an anti-birth campaign in the name of "population control". They engaged in all kinds of anti-family propaganda. That campaign didn't end until the 2000s - but by then the South Korean population had internalized it. https://www.prb.org/resources/did-south-koreas-population-policy-work-too-well/
"The share of men who identify as liberal has held fairly steady for almost 25 years...Meanwhile, among young women, liberalism has exploded. Young women today are much more liberal than young men" https://thehill.com/homenews/campaign/3675477-young-women-are-trending-liberal-young-men-are-not/
"Liberalizing prostitution leads to a significant decrease in rape rates, while prohibiting it leads to a significant increase." https://doi.org/10.1086/720583
Why race-faking is on the rise: In the 21st-century West, there are far more advantages to being black than being white. https://www.spiked-online.com/2023/03/02/why-race-faking-is-on-the-rise/
It's only a matter of time before the typical spam comment becomes more relevant and insightful than the average human commenter. Eventually, spam comments will surpass even the most ingenious intellectual achievements in all of human history.
This sounds crazy. But if you don't believe that the human brain is made of supernatural magic and maximally efficient at generating intelligence, which is very unlikely, then this is an inevitable outcome.