Instruction sets are just ways for us to allow it to do tasks faster. All those years using AI learning to get to facial recognition, voice recognition, patterning. We were creating shortcuts, a lot of them.

You don’t need instruction sets, it could all be done without them. It just makes this faster, need less overhead.

What we have been doing is creating those types of instructions and making them faster. What do you think instruction sets are? Task specific operations for a specific architecture.

We are living the cumulation of AI instruction sets to create an overlaying architecture. Learning to apply these different AI assisted learned skills into a more robust arch.

No it’s not really an arch, more like includes in C++. Eventually a lot of that work will become hardware. Think of encoding codecs, we started moving that to hardware and away from software.

The server farms are playing a huge role in this. The they are the uarch branches that currently are being formed. As we develop and merge these, eventually we will get to developing consciousness in some form.