The honest link is prediction
Conway's Life is one of the cleanest demonstrations of a hard prediction problem. You know the rule. You know the starting cells. Yet a pattern such as the R-pentomino is not obvious from inspection; you usually have to run it.
That idea sits close to Stephen Wolfram's discussion of computational irreducibility: for some systems, there may be no shortcut that predicts the outcome much faster than following the system itself.
Rule 110 made the argument sharper
A New Kind of Science spends serious attention on one-dimensional cellular automata, especially Rule 110. Wolfram presents Rule 110 as extremely simple at the rule level yet capable of universal computation when supplied with suitable initial conditions.
That does not make Rule 110 the same object as Life. The useful connection is narrower and stronger: both force readers to separate rule simplicity from behavior simplicity.
Why this belongs on a Life site
Many readers first meet cellular automata through Life, then later meet Rule 30, Rule 110, neural cellular automata, Lenia, or artificial-life systems. A research-oriented Life site should give that bridge a careful introduction.
The boundary matters. We should not claim that every complex system is explained by Life. We can claim that Life gives a concrete, playable case where local rules, emergence, and limits of prediction are visible on one board.
Working takeaway
Use Life as the first laboratory for simple rules and hard prediction. Then read Wolfram as a broader argument about why that pattern may not be an accident.