BY ML, you meant machine learning right? I did some homework to find out what is ML. Basically (at least as an idea) is not as fancy as skynet or what people think. As an example, the software (say netflix) categorizes movies base on various tags (horror, 90s, award-winning, etc) & when you watch movies, it plots your viewing history made a custom selection that best represents your preferences. Of course, it can be more than that, but that's the basic principle.In general, I'm hearing some really cool things about ML. Not just super resolution but there may be a surprise regarding backwards compatability soon. Kind of been hoping and somewhat expecting ML to have a big impact on older games. In the end, tools will be the story all gen. Microsoft's ML efficiencies will enable great possibilities with tools.
The problem with running ML real time and why we didn't see it applied much in gaming beyond some light implementation in BC on the One X is that it can be computationally heavy. Hardware accelerated ML just means you can run real time models and still have a lot left over. If you bought the Series X for hardware capabilities, you'll get what you paid for. DF freeze frame comparisons would've never justified the power delta regardless.
My curiosity is how does ML do in the context of improving gaming, say resolution....Anyone has any links how it might work?