Show me your data—not just your sleek pitch deck,
Your methods, your dictionary, the systems you’ve built.
Your coverage and demographics laid bare, not kept in check.
Show me your data—don’t hide behind proprietary walls.
Indulge me in your inclusion and exclusion criteria.
Publish the methods, share the code,
give the “how” behind the “whoa!”
Show me your data—a simple Table 1 would do,
and perhaps you could throw in a Figure 1 flowchart too.
What corpus trained your model?
What metric forestalls your errors?
How do you validate—internally or externally?
Show me your data—what does your missingness reveal?
How do you handle the gaps where information should be?
Can you acknowledge your biases with candour,
not conceal the limitations that shape your algorithmic prophecy?
If you truly are the best—as your marketing claims,
If you provide immense value and genuinely care about the populations you serve (yes, serve, not just aim to profit from),
Then, publishing research should feel fair.
Show me your data—a simple request, nothing grand.
Share methods, share learnings, build bridges of trust.
For all your AI promises and future so planned,
Until transparent beats opaque, “AI for good” is just a take.
Show me your data.
An ode to organizations building health apps, wearables, advanced analytics and AI. In a world where algorithms increasingly shape healthcare decisions. This is a call for transparency from those who build the tools that influence health and how care is delivered.
But Did You Live? 