I'm sorry to break the hype but honestly this looks sketchy as fuck to me. As in, a fake. OP has stated (I assume this is in good faith) that this comes from a 4chan thread and linked
this (original no longer available). The 4chan post is literally just:
"A deadman's program has activated and you have been selected as part of the reception avenue because the content is related to video games and you are marked as a media receipt."
So supposedly someone set up this deadman's system that, for some reason (you can verify this with a quick google search), ONLY targets 4chan and NO ONE else, not media agencies, not other websites, not civil rights organizations... literally just 4chan. If this was a literal dead man's switch then I'd expect personal details of the author to be published since, well, he should be dead, but there's nothing in the post.
Okay, now onto the slides. To begin with there are occasional grammatical errors, I'm pretty sure I saw an "it's" instead of "its", which is something you would definitely NOT see in a corporate type presentation. One of the first images at some point mentions:
alter the player's individual experience (psychological manipulation tactics) causes a consistent and dramatic increase recurrent revenue streams.
Besides the grammatical weirdness at the end, I strongly doubt that even the shadiest corp would use such explicit language to describe the concept. Anyone who's read any of EA's research bollocks or looked at some CIA/NSA leaks will know that more "political correct" or technical terms are used. Speaking of technical terms, for a presumably corporate presentation designed to convince executives or present a system, there are remarkably few actual pieces of data. No tables, no practical examples...
The "some real life data" section features some fairly weird things. To begin with, if this is a internal report then I'm not sure why the company's name is omitted. Then, there's this:
For example, when we track user cell phones, we can tell if they are driving past one of our partners billboards.
Again, grammar, but also that's not how phone tracking works. Unless they can somehow convince people to keep a GPS-enabled app open at all times, every other geolocation system is not nearly precise or fast enough to be able to do this. Anyone pitching an idea like this with a straight face would need to do far more than just writing:
We can monitor their reception to every single ad they pass by
which also sounds super janky and nothing like what you'd read in a meeting document or business pitch. Then there's this:
our partners (...) that have security cameras can opt in to our partner's services. They will send a stream of their store from their own security cameras to our servers where an AI...
To begin with no one with technical knowledge would write "an AI" like it's a common everyday app or something. Also, there are many technical roadblocks to streaming security camera video to a private server, such as the fact that most security cameras are CCTV (closed circuit), on different formats, not connected to the Internet, and of course privacy laws. You'd expect these to be addressed in such a revolutionary technology presentation but they aren't.
At any given moment your phone is broadcasting bluetooth, wifi beacons, and cell information to cell towers
This is not true (wifi and bluetooth can be disabled), "cell information" doesn't sound like what a technically-informed person or someone trying to sell an idea would write.
Our software takes in things like acceleration XYZ, geoposition, SMS send/receiving timings, Call send/receiving timings, GSM strength and estimated XYZ location...
Who the fuck writes "acceleration XYZ" to refer to acceleromter data? Same goes for the location. Regardless of this, the implication here is that an app like Facebook would be able to access this data. However, at the beginning they write:
Our software designed in Xamarin for all mobile devices
which implies that the software is its own app; Xamarin is an app development framework, look it up yourself; it's a front-end framework as well so it's not something you would "plug in" to an existing app like Facebook. This is contradictory, and who the fuck would install such an app anyways?
Maybe I'll continue later
Continued:
This is gold:
Use engine noises from our engine noise hashlib to detect type of car for income guessing purposes
In case you don't know, a hash is a mathematical function that converts an arbitrary piece of data into another data element (usually a string of characters or integer number) that has fixed length. There's no reason why comparing engine noises would require hashing them, let alone maintaining a library with hashes of them. Traditional hash functions also wouldn't work for sounds because the output of a traditional hash function can change dramatically even with a slight variation in the input. There is "fuzzy" hashing which can work in theory, but it's a fairly novel technique and you'd expect this to be mentioned somewhere.
The AI begins a new testing lifecycle that starts when the game session closes. It will patiently lie in wait for the high value distraction event to end
"It will patiently lie in wait", come on. It sounds like what a 10-year-old thinks AI is after watching Transcendence. Also, there's no such thing as "AI lifecycle" unless maybe in evolutionary algorithms, but those aren't mentioned anywhere.
It discovered a correlation between the voice pitch adjustment away from the normal standard deviation, and that women would buy more in those 48 hours
There is no such thing as "normal standard deviation". Normal distributions and standard deviation do exist in statistics, so maybe this (incredibly incompetent) researcher was referring to the standard deviation of a normal distribution? Even so, without more precise context all of this is meaningless; the slide has no context. Also, standard deviation is a measure, and as such it should be accompanied by some kind of value.
For example the recent case outlined in schedule "P" shows how developers or persons targeting children for emotional manipulation is illegal by the CJEU ruling P(1).
CJEU is the European Union's Court of Justice. But conveniently, "schedule P" and the supposedly related ruling are absent, so this supposedly real fact that the text refers to is entirely unsourced. So far googling around I have found no such ruling either.
Let's take a look at the paper documents. I'll just quote this:
Side-Channel Data: Data not gathered via a data point in itself but of the implementation of the data.
This sentence makes no sense. No one would ever write this on any technical paper.
The password submitted is a secret, but the MD5 hash thereof can be smart-metasearched before it is salted along with other details of the user.
You can't do something with a hash "before" it is salted, salting a hash function just means adding some extra data (that isn't the password being hashed) to the input. There's no "before" or "after", a hash function is either salted when it is calculated or it isn't.
Credit to /u/
-The_Blazer-