Chess will never be solved, here's why

Because the more piece tablebases we can finish, the easier it will be to determine what forcing lines in the middlegame and even late opening can lead to those types of endgames. It's not that we need a 32 piece tablebase, it's that we need a partial solved chess database expansive enough to cover the types of middlegames and endgames that the already-optimized opening books can lead to. The middlegame is the last link, but it could greatly reduce the number of tablebase pieces count needed.

just let me know/message me if tygxc comes back and starts spreading disinformation again - bye

What are you, batman?

Do it for Gotham!

Another aspect is implementing tablebase algorithms that understand universal piece canceling-out. For example, Queen vs Rook, what specific pawn ratios on each side in addition to the queen and rook always make it a perfect-play win for the queen or a draw. If universal ratios like this could be discovered for more complex endgames, like rook + pawns vs knight + pawns or KRKN + x number of pawns on each side, this would "solve alot of normal looking endgames as the result would be based on which pawn ratios always will simplify down to winning/drawn piece tablebase positions. Alot of 12+ total piece endgames are the sum of an already solved pawnless endgame (like rook vs bishop) and a ton of pawns. If a set of say, 100-200 rules could be programmed in to know when the quantity/file positions/and spread of the pawns on both sides trigger the piece endgame to be a win/draw/loss, this could be combined with the piece-only tablebases to evaluate if positions with a ton more pieces (pawns) are also always a certain result. Current tablebases already do this within simplifying from pawnless endgames down to smaller pawnless endgame, via "distance to conversion" calculations. If for example, it was ever discovered that being 2 pawns up in rooks and pawns endgames, then you could add a few dozen more specific heuristics, such as whether pawns are doubles or not, how advanced they are on each side, whether the pawns on each side will clash or if they are each majorities on opposite dies of the board, and various combinations of these parameters. Then the computers could figure out what conversions to simple piece-only endgames are necessary, based on the pawn parameters. This could allow finding "perfect play" moves in positions with pieces that exceed normal tablebase limits.

Somebody showed me a PGN of the longest forced mate that has been discovered. It exceeds the 50-move rule without a pawn advance or capture... by about 500 moves. I wonder how it can be verified, though? What program would be able to prove that it really is a forced mate in 549 moves?

That's some KQNKRBN position.

An even longer mate was found within 8 pieces but it involves 2 bishops of the same color. Ironically alot of the solved tablebases are totally unrealistic endgames, like how many "perfectly played" games are going to end up in an endgame with a queen and knight vs 3 other pieces and no pawns for either side, let alone 2 bishops of the same color and a pawn+rook vs a queen and pawn. That's why ironically even the most complex solved tablebases don't help much with solving chess from the starting position.

I hope whoever discovers the solution to chess gets fairly compensated for their major contribution to humanity

This is like asking how you can prove geometry theorems.

Such positions are taken from tablebases. The 7-man tablebases are built on the 6-man tablebases, which are built on 5-man tablebases, etc. successively. Now take a 3-man tablebase...king vs, king and queen for example, and would you ask how you can "prove" that K+Q mates a king? Tablebases are exhaustive and built on trial-and-error brute force calculation.

You cannot personally prove it to yourself, not in your lifetime anyway...nevertheless, it is 100% accurate.

I guess there's a point where you have to rely on the integrity of the computing machine

It's more likely the "perfect" game of chess whatever we ponder that to be, ends up simplifying down to either a winning or drawn KPK position or maybe something like rook vs 1 other pieces. The fascinating tablebases like 2 knights vs pawn, minor pieces vs multiple pawns, Queen vs a ton of other pieces...etc are usually the result of highly imbalanced games that were likely less than perfect with tons of mistakes on both sides.

There also a point where you must rely on experts, because no one person can master everything in their lifetime. So when professionals with credentials, degrees, etc. are in consensus about something, they know far more about it than YouTubers, parents, Joe next door, etc.

This is applicable in all sorts of subjects...

What if there is a question about which none of the experts have relevant expertise?

If you are seeking expertise, then it behooves you to seek the relevant expertise. If you think it doesn't exist, then you either need to overcome your bias, research better, or if you truly find something unknown, pull an Einstein yourself.

'if you think it doesn't exist' could be a 'motivated think' by whoever not Wanting to think/know that the expertise not only exists but is well established too.

First you say chess has been solved, then you describe the act of solving chess and how that still has to be done

A familiar pattern!

What championship contender played Go to train?