Guy's really onto something here.
'#DataDriven' contains a bunch of assumptions worth challenging, not least that #data is not #truth - it is only what you #choose or are able to #measure.
And, as well as being a way to avoid thinking - just set your #OKRs (or increasingly...
'#DataDriven' contains a bunch of assumptions worth challenging, not least that #data is not #truth - it is only what you #choose or are able to #measure.
And, as well as being a way to avoid thinking - just set your #OKRs (or increasingly...
https://twitter.com/guy_herbert/status/1306833947561734146
...pick your training dataset & model) and iterate until something breaks - '#DataDriven' is an insidious way to avoid #accountability: "We're just doing what the data says!"
Of course, the flipside of that is there is then no excuse NOT to be #transparent; why hide the data...
Of course, the flipside of that is there is then no excuse NOT to be #transparent; why hide the data...
...if your contention is that it represents 'the #truth'?
Maybe Mr Cummings et al. should beware what they wish for, as - far from building #trust - systematically concealing #evidence tends to have the opposite effect.
If you think about it.
Maybe Mr Cummings et al. should beware what they wish for, as - far from building #trust - systematically concealing #evidence tends to have the opposite effect.
If you think about it.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
