James' talk reminded me of many things including
- the challenge of interpreting genomic "evidence" in making a clinical diagnosis - a topic at the heart of the recent NHMRC workshop towards principles for translating omics based research into clinical practice
- the ease with which p-values and null hypothesis significance testing can be misinterpreted - an subject given an excellent exposition by Professors John Carlin  and Gordon Smyth  at FOAM 2014
- the role of television (e.g., CSI) in shaping beliefs about what science can do.
...that last point can be described as the "back-of-the-briefcase effect", a term coined by Mike Brown at CVAB 2013 as he recounted dialogue from some spy movie along the lines of "OK, zoom in on that briefcase... increase image resolution... close in on that some more... right, now show me the back of the briefcase."
...which makes me think of the dual-edged sword of biomolecular visualisation (though "back-of-the-nucleosome" is a little less catchy).
And as this train of thought pulls into the station for lunch, I return to the beginning by commending another of James Curran's presentations to you: Is Forensic Science the last bastion of resistance against Statistics?