Two more takeaways from an excellent episode of the Ezra Klein podcast.

(continued from yesterday's post—link in the top comment)

1) Don't outsource your value prioritization to the system.

When we use social media platforms, we allow the system to tell us what to value.

(This is especially dangerous when the system is run by a for-profit company.)

The system's values are built into the user interface itself:

Our only options on most platforms are "like," "share," or "comment."

And try as we might, it's almost impossible to write a post and not care how many "likes" it gets. That's what the system is designed to make us want.

But could there be alternative models?

Platform designers could decide to incentivize other things. For example:

• How many people answer a question you ask;

• How many people disagree with your point and offer an interesting counterpoint;

• How many people agree to make a commitment that you suggest.

Questions: (1) What other models come to mind for you? (2) With the current social platforms, do you have any tips for encouraging actual responses from readers rather than just passive "likes"?

2) "Likes" and aggregate scores are ambiguous.

Not all "likes" are equal.

When someone "likes" your post, does it indicate "eh, cool post"?

Or "oh my god, this post changed my life"?

Similarly, a movie with a 50% score on Rotten Tomatoes is ambiguous:

It could mean 100% of people thought it was meh.

Or it could mean that 50% loved it and 50% hated it. (This is often an indicator of truly great art.)

Still, it can be overwhelming to find a movie you'll like without quantitative data.

There are a whole lot of films out there, and the paradox of choice is real.

That's where large data sets can be useful sometimes (as long as the system is carefully designed to prioritize the right values).

Tip: Instead of relying on general review aggregators, I love criticker.com. You rate movies you've seen, then it finds other people with similar taste and uses only their ratings to predict how you'd rate movies you haven't seen yet.