I just meant that causal models using algorithmic conditionals look a lot like the “timeless decision diagrams” that were roughly sketched in the TDT draft paper. It seems to me that both generate “non physical” correlations by way of links representing logical dependencies between physical processes, though that’s not quite the terminology Janzing and Schölkopf use.
I’m not seeing how this is coming close to rediscovering TDT. Could you explain?
I just meant that causal models using algorithmic conditionals look a lot like the “timeless decision diagrams” that were roughly sketched in the TDT draft paper. It seems to me that both generate “non physical” correlations by way of links representing logical dependencies between physical processes, though that’s not quite the terminology Janzing and Schölkopf use.