Hot ice cream was a thing that was done a few times during the sciencecy food craze in the early 2010s (“molecular gastronomy”, the most mastubatory name of all time).
It’s remarkably easy to do. There are gelling compounds like methylcellulose that come in a host of variants like F50, which makes a semi firm gel between 62-68C, or a4c which makes a firm gel between 38-44C, etc. As a result you can create an ice cream base fairly traditionally with 0.1-3% methocellulose dispersed (more for firmer, though it can become “chewy” like gelatin if you go too far).
Then to set you can heat a water bath (or simply boil water) and prepare it much in the way one would prepare dumplings. Spoon in the base and it will quickly set as it comes to temp that you can immediately serve. They are the same base of “ice cream” but hot, and they “melt” as they come to room temperature.
This technique has existed since at least 2013. It’s fairly lackluster unless you adjust the ice cream base significantly to work with a hot preparation. Further, it appears she’s doing something mediocre here that’s more suitable for mass production and more akin to a warm milk shake, which is basically just melted ice cream? That’s easily possible with modern ice cream loaded with stabilizers but that compromises texture (which is another problem with the methylcellulose approach: if you don’t absolutely nail the percentage it will taste like a gel and not an ice cream)