What's a good career for someone good at recognizing patterns, but not a math whiz?
![[Post New]](/jforum/templates/default/images/icon_minipost_new.gif){kind=icon}
Anonymous
PP. It's not the raw capability of the computers that I quibble with, it's how much it costs to gather the material and get it correctly ingested. And to program the computers to correctly interpret/translate. Right now I'm working with three non-standard antique languages for my hobby genealogy. Not even sure who might have or load an English translation dictionary for the ancient form of one of them. There is a lack of standardization of old physical materials. And the cost of digitizing is fairly prohibitive. The Mormons do a lot but their online FamilySearch engines are crap right now. It's taken 10 years to get from digitized microfilm to search engines that only find some relevant (mostly incorrectly transliterated) crowdsourced names. Plus when amateurs get involved they put provably wrong info out for the AI to scrape and proliferate. And the AI does. Ancestry "Hints" anyone? An example of an amazing specialized pattern recognition computer program is the one that recently won the challenge to find words in the charred scrolls from a burned Roman villa library. Quite a lot of work has gone into that and so far we only have a few new words. After it gets going, how will those words be interpreted and set in context? By a person with Classics training. I have a 30 year background working at jobs that involved imaging, digitization, and search algorithms. There just isn't infinite $$ to make all relevant info scrapeable by bots. Never mind the AI hallucinations. Just comparing Microsoft's corporate search products to Google is enough to make one throw up one's hands. If it's so easy to computerize knowledge generation, why is Bing still so bad? There will be plenty of room in the future for humans to use computerized tools to examine their pattern recognition hypotheses. |
Anonymous
| Medicine. |