Not really. Even the people who train the models and code the algorithms don't have a lot of control over how they respond to ever-changing inputs - and that's kind of the real problem IMO. There was a recent thing on Twitter, where someone discovered that it would consistently generate thumbnails focused on white faces when both white and black faces were present. At a gross level this is clearly because of inadequacies in how that algorithm worked, but literally nobody could say exactly what was wrong or how to fix it (especially without introducing new kinds of bias). Facebook has the same kind of problem, at approximately 10x Twitter's scale.

It's a machine that nobody knows how it works now, and therefore nobody knows how to fix it. Mark can influence it in very vague and general kind of ways, I suppose, but there is no fine control and often these gross adjustments bring their own unintended consequences.

If we don't understand how something works and it seems to be causing negative effects perhaps we shouldn't be using said algorithms in the largest social network.

I'm not actually going to disagree with that. It's a perfectly valid point of view. However, I will also add this thought, which I also expressed internally.

Neutrality doesn't just happen. No matter how neutral a system it is at one point in time, people will continue gaming it and it will cease to be neutral. There must be continuous active response to maintain that neutrality. This is a lesson Google learned with SEO long ago. Passivity is not the same as neutrality, and insisting on one often makes the other impossible.

That's not meant as a refutation of your point, but perhaps it's worth some time to think about how a "no algorithm" version of Facebook would actually play out. Here's your starting point: Twitter but 10x as large.

> Neutrality doesn't just happen. No matter how neutral a system it is at one point in time, people will continue gaming it and it will cease to be neutral. There must be continuous active response to maintain that neutrality.

This is a scary thought - an algorithm that "decides" what neutral is, and pushes all its users towards its predetermined neutrality.

All of this talk is nonsense if we can't define neutral - which we can't. If Facebook is going to be an arbiter of that, they need to be responsible for the content because they choose what content is shown and when - if it was just a timeline feed, they would bear no responsibility.

Facebook wants to make all the money and not scare off customers - a clear conflict of interest.

> if it was just a timeline feed, they would bear no responsibility.

Did that work for Twitter? They've hardly been immune from criticism either, despite being ~1/10 the size of Facebook and being favored by the ruling party in the US. Chronological timelines aren't silver bullets. They're only chronological among that which is shared and there are still huge differences in how much content gets shared/like/retweeted/whatever. Those are the differences that people will exploit, leaving things only "neutral" in a narrow technical sense that has nothing to do with effect. It's naive in the extreme to think they actually solve the problem.

I think you are quite misunderstanding the point.

My point isn't that a chronological timeline fixes the problem of social media. My point is that if you are going to decide what is shown, you are making a choice, and should be responsible for that choice. It is no longer just the user's post - because you control its avenue, you are responsible for it and how it spreads, and becomes "your" speech, in this case, the algorithm's, and its company's.

So you're suggesting that Facebook should try to absolve itself of responsibility, but not actually solve the problem? No offense, but that seems exactly opposite to where you started. We're done.

It causes positive effects in the form of higher profits for the company.

Wow!

That's really an argument someone is prepared to make.

One role a CEO can play is to direct a company to act in particular ways.

The responsibility for anything Facebook does or doesn't do absolutely lands squarely on the companies leadership.

Saying something like: 'welp, it's too complicated to understand, but fuck me look at all this money and power...

Isn't going to convince me.

I think you're missing my point. I'm not saying that Facebook should give up. Everyone there should make every effort to improve. My point in bringing up complexity is that it's going to take a lot more than Mark waving his hand and magically making everything better. This isn't some video game where one player hits a button and thousands of minions instantly rearrange themselves and all of their actions are resolved in milliseconds. Anyone who knows anything about complex systems - especially those involving people - knows that's not how they are.

In reality, no matter what leadership says or does, actual change will still require continuous effort from literally thousands of engineers, data scientists, and others. It will take time, as all such things do. Petulantly demanding that things happen faster than they can happen isn't going to make it so.