AI: NYT piece on automating finance jobs using AI: “Some of Wall Street’s major banks are asking the same question, as they test A.I. tools that can largely replace their armies of analysts by performing in seconds the work that now takes hours, or a whole weekend.”
>ChatGPT-4 beats humans in test of social intelligence: “ChatGPT-4 exceeded 100% of all the psychologists, and Bing outperformed 50% of PhD holders and 90% of bachelor’s holders.”
Not really. Check the paper. The test is some random scale from an unpublished 1998 Phd thesis from Iraq, written by the first author. Not even clear what this scale measures, and no detailed metrics are given, or examples. Presumably, AIs just read this or related content in their training, so they are just memorizing these results.
>ChatGPT-4 beats humans in test of social intelligence: “ChatGPT-4 exceeded 100% of all the psychologists, and Bing outperformed 50% of PhD holders and 90% of bachelor’s holders.”
Not really. Check the paper. The test is some random scale from an unpublished 1998 Phd thesis from Iraq, written by the first author. Not even clear what this scale measures, and no detailed metrics are given, or examples. Presumably, AIs just read this or related content in their training, so they are just memorizing these results.
This is an Arabic copy of an old English language test, from the 1940s. https://ehive.com/collections/5889/objects/534738/social-intelligence-test-george-washington-university-series Looks like a kind of unusual IQ test.