It's a bit abusive of the Harvard brand to throw the name in a post just so you can get some training data for model. At least include a writeup of what you're doing so people who click through get something out of it too.

looks like he's doing a machine learning to determine what images are aesthetically pleasing to people?

or if the order of images determine how much more likely it would be liked.

but yeah, abusive(???) in using a name with no explanation of relationship, no explanation of purpose after participation, etc.

then what'd be the point of having the harvard brand at all?

Exactly...isn't that what you're paying for?

Hi guys, I'm running an experiment for one of my courses and need to collect some data. I would really appreciate your help.

Thanks!

happy to help, but some instructional text on the screen besides 'machine learning experiment' would be nice.

actually, upon thinking about this more, i'm somewhat skeptical that the dataset you obtain from this will be useful..

Thanks for your help! I hope to find some interesting data, it is an experiment. We'll see what the results yield.

Wow, these pictures are really cool! What is this experiment all about?

The photos are credited to Trey Ratcliff, http://www.stuckincustoms.com/

I don't want to bias the results. But once the experiment is complete I will be posting the findings