The intent of this to-be proposal is to begin distributing the main Polkadot Treasury into other assets than DOT. It will convert 1,700,000 DOT into USDT gradually over the next year. Depending on the staking rate, this accounts for about 13% of the new DOT that would be minted into the Treasury as a result of inflation (without even considering the existing Treasury balance).
With USDT in the Treasury and the latest updates to the Treasury pallet, users will be able to make Treasury proposals directly in USDT. They can also add a payout schedule, for example "X USDT upfront, and then Y USDT each quarter for a year", which would be optimistically claimable (but cancellable by governance should the beneficiary not deliver along the way). The stability of a, well, stablecoin and scheduled payouts should enable more milestone-based work and alleviate the need of teams to navigate the balance between asking for a single upfront reward based on trust vs. the time it takes to advocate for many smaller proposals.
Further work remains to be done on a more strategic level, like an ideal Treasury allocation and the mechanisms to achieve and maintain it, but I think 13% of new issuance into USDT is relatively uncontroversial, leaves plenty of room for more options, and is a good start.
On a technical level, this uses the same mechanism as referendum 231, which used (is still using) HydraDX to convert DOT to USDT over a prolonged period of time. Instead of depositing the USDT into an account dedicated to Fellowship salaries, it will instead deposit the funds into the Treasury's main account on Asset Hub (14xmwinmCEz6oRrFdczHKqHgWNMiCysE2KrA4jXXAAM1Eogk ).
See the full proposal doc from HydraDX for more info.