Very annoying one (I'm not good at cryptic crosswords) because I knew the answer but didn't know why. Grok explained the reasoning even if I still don't think it was a good clue (although it went through all the steps I did to arrive at it). Not an endorsement of Grok, but stating that it was the one that got it out of the three public AIs I tested. I used less calories.
Prompt:
Explain the cryptic crossword puzzle clue:
Started eating odd bits of Brie with Spanish Red (3,4)
https://grok.com/share/bGVnYWN5_82a144ca-ca01-4a2d-833c-18ade6c017d0
Wordle 1,348 4/6* (this could have easily been a 5 or 6. I rotate starting words.)
🟨🟨⬛⬛⬛
⬛⬛🟩🟩🟩
⬛⬛🟩🟩🟩
🟩🟩🟩🟩🟩
This game is an interesting one regarding how to optimize guesses, tempered by the frequencies/positions of English letters in words.
Daily Word Score:
wordglyph.xyz/508
🧠 9 ②②①①②①
Wordle 1,333 2/6* (yesterday was a quite rare 5, but the two day average is now 3.5 so 2s have a use occasionally). An honourary 3.5.
🟨⬛⬛🟩⬛
🟩🟩🟩🟩🟩
Current grain of choice - extra coarse bulgar (1:2 ratio of liquid, with roughly the same protein ratio as wheat). Picture set to medium quality because it's not artsy.
My guess is AI and technology, for many, will have, at worst, a neutral effect on memory. My reasoning is that spelling autocorrection has reduced the rate at which I mispell words rather than made me more likely to mispell words and I'm good at remembering a route even if I used a map the first time. Where, speaking personally, there is a danger is that curiosity makes it very easy distract myself and there is a danger of not focussing on things in-depth, which is covered later in that article. I don't confuse a Perplexity.ai (which I have found useful) explanation of something with in-depth knowledge and I sometimes hit the limit of an AI's knowledge - some of them are crap when that happens. The most annoying one is inferring what people usually mean when what people usually mean is absolutely not what I'm asking. It's not that I gave it an idiotic prompt it's just that it's more likely people make a mistake so it assumes I'd made a similar mistake. Prompting the man computer symbiosis equivalent of Basil Fawlty giving his Mini Metro a damn good thrashing. Plus a lot of stuff in the training sets is as likely to be wrong as the people who initially produced it.