You'd be wasting your time. Your data is versioned. People smarter than you have long thought about this.

It doesn’t matter. This information has a half-life. What I “liked” diminishes in value over time. What advertisers care what I clicked on 1, 5, 10 years ago?

Imagine that at different points in time, the confidence in your persona can be quantified. Say at some point in time last year they were 89% confident of the type of person you were, but today after months of using a ton of random fake likes and clicks they are only 16% sure.

They could choose to ignore low confidence measurements and still assume you are the same person from that point in time last year. Given that they can measure the standard rate of change in people’s interests (per interest even), and how quickly interests fall out of style, they can then extrapolate how relevant certain ads might be for you today based on your old data.

Imagine that. They don’t just own your data today, they own all the forecasts of your data for years to come. Go home and sleep on it.

What if the person who gleefully shared everything on social media, and the person who hides their trail actually are different? People change their beliefs and spending habits all the time. It why we look back on our own writing from years ago and cringe.

The best solution is to just not use Facebook.

Better if you can get your account suspended for something that doesn't provoke a law enforcement response.

Agreed, if you don’t want them to have data on you the best case is to never use anything FB owned, including Instagram and WhatsApp.

If you already use FB services, stop using them and don’t give anymore data. They will still know about you, but at least you won’t give them any new information.

But do they actually do that? It seems like a lot of effort to market to someone who's less likely to be successfully marketed to.

Unless you are mentally disabled in some way, anyone can be successfully marketed. People like to believe they are special snowflakes that can rise above the influence of marketing, but hit them with the right marketing message and you got them.

> People like to believe they are special snowflakes that can rise above the influence of marketing, but hit them with the right marketing message and you got them.

Maybe so, but the success of marketing has different probabilities for different individuals. So, if you're Facebook, it wouldn't make sense to invest in de-poisoning poisoned tracking data if they'd have a low chance of success with their marketing messages. Also, I'm doubtful any de-poisoning technique is perfect, so the resulting profile is probably worse for targeted marketing, all else being equal, as well.

Right, that’s why the only way to win is not to play, there is no safe level of exposure to Facebook or Google.

Sure, for forensics, but how would FB's AI know that your new data is garbage?

Isn't that like NP-hard or something?

Furthermore, would they care? Realistically, you're padding their numbers. Things can be popular simply because they're liked (whether it's synthetic or not).

Depends on how you put in the data.

If you do it all at once it’s trivially easy. They know a person doesn’t just become a new person overnight.

If you commit to creating an elaborate fake personality by doing things your fake self would do, it can be harder, but not impossible. Certainly not a problem for some of the greatest minds in Silicon Valley.

Meh...the whole "Silicon Valley genius" trope is another myth that needs to die. Being able to code a depth first search doesn't make anyone a 'genius' and there are limits to what AI can do even with bushel baskets full of data.

For many years Facebook thought I was Jewish and was spamming me with ads to become a rabbi and move to Israel, but I am the least Jewish person you'd ever meet. And that could easily be determined by publicly searchable data about myself.

I've faked my phone number for about two weeks on Android when I installed their app to try it. 3 years after, they still ask me if my number is some random Italian number, I'm not sure they have that many checks in place.

What you aim is not to remove your data, you want Facebook to think you are a different person than you are by adding false data in your account.

You will not be a different person, you will be “Person who thinks they can fool us into being a different person by feeding in fake data.”

You’ll probably get ads for people who are paranoid or security conscious.

It is a great idea, and I think it can be pulled off without being (easily) detected as decoy activity by making occasional submissions, than like 20 submissions a minute. A very simple way to do it on top of my head:

- Seed a crawler with your own profile (or any other). Make it find 50~100 public contacts with public timeline in your network. Scrape their shares along with their timestamp deltas to their corresponding previous shares. Omit the very first, for that it lacks a proper timestamp delta. Put them all in a collection.

- Pick any one by random, post it all the same, after the same amount of delay as the original share against its previous share. Repeat.

You may shuffle the sentences whenever there are more than just one sentence, to obfuscate things a little.

What you suggest is the only thing that will succeed in preserving privacy. You cannot hide your data as they can still track you even if you do not have an account. We have to lie plausibly to contaminate the data. The important thing is that the data should be carefully generated not to be flagged as synthetic.

Ublock + NoScript + self-destructing cookies makes one much less convenient to track. Your IP and browser headers are all that's left.

You are right but the problem is this: I would like to use Google maps but I do not want Google to have a perfect profile of me so I try to add plausible false data.

I love this idea. Maybe someone could set up a service where you enter your fb password or upload your fb cookie. Then it would just destroy your profile with data.

How would it handle historical data? I don't think you can backdate messages and photos.

It'd look like automated spam, and they'd block/flag it on the spot.

What would happen to your fb account? Do they only block the account or the spam?

Also could you use ML to be more discrete to avoid the spam flag?

Not sure about ML, but there're several problems to solve before that. The service would need IP address that aren't from known cloud provier, otherwise it's trivial for them to notice and block. Also it'd have to look like a real user, so a full browser (posting directly to their endpoints will get you flagged). Then, each user can't post too much and the IP would need to be recycled as well.

Too much trouble to even start that.

You can rent a botnet of desktops from residential IP.

For example, the Hola "unblocker" C&C people have this service. https://luminati.io/

A service that continually updates your profile data (well the non-invariant type), adds/removes random "friends" and the like would be interesting.

Do it progressively enough, and when would FB know the difference?

Something like this? - https://likenoise.com/

Maybe? But with the influx of new activity, it would be reasonably easy to see the line of real activity vs fake activity is on a timeline.