What if a crossword puzzle evolved in real time? All correct letters stay locked, but unsolved words and clues constantly change.

Stuck on a word? Don’t worry — eventually it’ll change into something you can solve. The puzzle adapts around your progress, keeping things fresh while preventing dead ends.

It’s like a "living crossword" that adjusts itself to your solving speed and skill.

Would you try such a crossword?

I used to like these in old magazines that would have multiple clue sets for the same puzzle layout. Try the hard one, then do the easy one. The mechanic in an interactive context strikes me as challenging. Is it just time? Is there a hint submechanic? Is there a points mechanic? Is there a difficulty labeling (like give me the Saturday version first, then switch to Wednesday). How much cognitive load to add to the game? I'm also a little skeptical that the (implied) llm generation of clues or hints could be implemented without frustration. It seems like an interesting idea to explore but many paths in that game design tree seem like they could be frustrating rather than fun.

This would actually be rather neat because whenever I do crosswords, I find that I can do a certain % of it, and then basically get stuck.

Being able to reshuffle the rest would be interesting.

Nope. I don't want my puzzles to dumb themselves down for me. The whole point of a puzzle for me is the challenge. If they're too easy they're no fun.

>Would you try such a crossword?

As a long time cruciverbalist (40+ years), I would hate it.

Not because it's interactive, but because it adapts to the solver. Which kind of defeats the purpose of doing a crossword.

Which is to stretch your understanding of language and vocabulary. Adapting to "your solving speed and skill," is dumbing down the crossword -- which is the opposite of that.

Yuck!