tokyokevin: Kevin Ryan in Tokyo

Kevin putting it out there

  • Women and Men in University

    In this newsletter (blog post) Celeste Davis looks at the reasons why fewer men are going to college in the US.

    For every 1% increase in the proportion of women in the student body, 1.7 fewer men applied. One more woman applying was a greater deterrent than $1000 in extra tuition!

    This is also reflected in certain “Male” jobs and “Women’s” jobs. With the new political weather in the US, it will be interesting to see if this trend slows or accelerates. I’m guessing on the second.

    I’m hoping women will not give up and continue to “take over” more universities and departments and whole careers and career paths.

  • Reasoning and/or AI?

    Mike Caulfield blogs about reason and reasoning and surprisingly doesn’t care how AI got there, and who is still doing the work. Google and Perplexity have launched Deep Research engines, AI help for regular research. He talks about SOK (State of Knowledge) and the next step up SSOK (Statements about SOK), which often include references to the knowledge (links or quotes from other work).

    “A question sometimes comes up at this point — well, look at all the work you had to do when you verified this. Isn’t it actually you who is doing the reasoning?

    This question doesn’t really understand cognition well. In a sense, it’s always me doing the reasoning.”

    Is it really research if the cognitive process is being somewhat replaces with AI? I think it is. It just becomes quicker, and more a filter than an engine itself. Worth thinking about.

  • Cool Biomass Visualization

    Multi-step explanation of how much living stuff is on earth (biomass) and then compares that to man-made stuff (technomass). Eye opening. Via Kottke.

  • Al is making us stupid

    Research by Microsoft (pdf) shows tech industry users show a reduction in critical thinking as use of AI increases.

  • Fake Research

    The Whole Earth Catalog was vetted. Not so for many journals.

    Is caused by an evaluation process much like Google’s PageRank, where a paper is rated by how many links there are to and from it. That’s how over 400,000 research articles in the last 20 years are probably fake, created by Paper Mills. From Nature.