This has the potential to change computing forever. If it actually works, Neuralink could be the pioneer of a completely new computing method.
It would, if it works, one day.
This is (partly?) learned behavior. Children under the age of 5-6 will have audible self-directed speech, but eventually they learn to mute this and remove vocalization altogether.
These "augments" will likely be, at best, some kind of subsystem you can tap into, like an extra limb, or a memory you didn't know you had until you "recall" it.
Now if it is the same thought process as controlling a limb, well that sounds like the same mental energy needed to type out something on a keyboard or speaking commands, both of which happen at close to the same speed as thinking the words to begin with (well, talking definitely, typing within a high enough fraction for most proficient typists).
Somehow it still ends with pop-ups and dark patterns.
Stop hyperventilating over this. There are serious research projects worth following that are far more likely to deliver useful technologies.