[disgruntledphd2]

> Optimizing for "engagement" requires that you offer the mind an endless set of distractions in lieu of good engaging content. Over time the mind starts to degrade to a point where its difficult to digest long-form content and enjoy it as a much healthier habit.

This is another really strong claim. I find it odd that most people are so sceptical of the claims established by social science research, except when it confirms a previously held theory.

TV optimises for engagement. Do you think that TV lead to the destruction of people's ability to engage meaningfully with the world, or is that distinction reserved for products invented after (I assume) you were born.

Cos I remember how much everyone gave out about TV (and I was one of those people!) and to see the entire argument wholesale lifted and applied to a completely different medium of communication is disconcerting, to say the least.

[luckycharms810]

Your TV (at least in the past ) could not collect the following information:

- What ads did you stay for ?

- What ads did you listen to, or turn the volume up for ?

- What ads did you hear and see 50% of with 50% of the pixels being on screen at the time you heard them?

- How many times did you see an ad ?

- Whats the average times that an ad is seen before you might convert in an offhand channel ( including offline credit card purchases )?

- What ads changed your scroll velocity - if only for a few 100 milliseconds ?

- What content changed your scroll velocity ?

- Where were you when spent less time on your phone ?

- Where were you when you were more susceptible to a type of ad ?

These aren't signals that you might yourself notice, but these are being collected. They are being collected on almost every website you visit, and an incredibly large data set is being collected. This training set is for you and you alone, but similar training sets can be found.

This training set is now used to determine what content to show you next so that you can stick around for another ad.

This is 'optimizing for engagement'. We are well past the stage of "It's the superbowl so lets play an Axe commercial'.

Some of this sounds fairly hyperbolic to folks who haven't spent time in the industry - but I promise you this is very commonplace non-controversial data collection in the adtech space.

[disgruntledphd2]

> Some of this sounds fairly hyperbolic to folks who haven't spent time in the industry - but I promise you this is very commonplace non-controversial data collection in the adtech space.

I spent well over half a decade in adtech, and while I don't dispute that this data is collected, I definitely don't think that it's particularly useful.

Fundamentally, the best information to use to predict what you'll buy is what you looked on on a site/app and what you've bought in the past.

The rest of the stuff you list out has really, really small effect sizes, it does make a difference if you're Facebook or Google (though the effects are still just as small) but given their scale, it's meaningful revenue.

Additionally, all of this data is incredibly sparse, so it may help at the margin, but definitely doesn't help overall. That's even assuming that the data can be reliably linked to an individual, which in most cases is definitely not true.

I think that you appear to be concerned more about the tracking than the marketing itself, is that a fair assessment of your position?