then go ahead and improve it, nobody is telling anyone to stop improving shit. but claiming you could use a tech at a point in time where it is not much more than a gimmick, people wont take it serious or too serious. just like with ChatGPT: just because it can return highly probable language doesnt mean its correct.

just because you can drive one "artificial tree" around doesnt mean you should pursue in making those on a scale fit to dent the problem, because in making those "trees" you just release more dead-tree-gas and make the problem worse before you can make it better.