Was This Lost Computer Code Worth Billions?
Vložit
- čas přidán 25. 12. 2020
- Jan Sloot claimed to have developed a revolutionary means of data storage which could reduce a feature film to just 8 kilobytes of data. But when an unexpected tragedy occurs, his investors are left scrambling as they try to recreate his work.
Music by Timmozoma: / timmozoma
Please consider signing up for our Patreon: / anomalydocumentaries
all sources:
web.archive.org/web/201606170...
web.archive.org/web/201502271...
steemit.com/blog/@tweakman/th...
jansloot.telcomsoft.nl/Source...
archive.org/details/valleyboy...
usbeketrica.com/fr/article/ja...
www.gids.nl/techno/sloot-twijf...
www.gids.nl/techno/jan-sloot.html
www.mithileshjoshi.com/2019/0...
jansloot.telcomsoft.nl/Source...
blogs.sap.com/2005/05/02/size...
jansloot.telcomsoft.nl/Source...
www.physicsforums.com/threads...
jansloot.telcomsoft.nl/Source...
This guy wasn't a misunderstood visionary. He was a straight up fraud. The stuff he claimed was literally impossible as evidenced by the fact that no one actually saw it and he was intensely paranoid about anyone peering under the hood. The only amazing thing about his story is that he somehow got an audience with a number of prominent tech people. I guess it pays to have a good lawyer.
The problem you need a little bit understanding of statistics/stochastics and compression algorithms to understand why. An average investor even in IT in that era didn't.
Information entropy would be enough to prove that the idea was impossible. When you measure maximal compression and measure the minimum possible entropy one would realize the claim is impossible. A sketch of a proof would be that the minimum number of characters encoded in an 8 KB string is strictly less than the minimum number of characters required to encode a full feature length film at max compression.
the only thing that is possible right now is 64kb and 4kb demoscene but then again those stuff requires a lot of 3d api from the OS so technically its only instructions of graphical animations. tho still amazing to watch how codes can squeeze animations from these compression algo that most of the time they can even fit a game like 96kb
another is john hutchison. said he could prove free energy was possible. made all sorts of claims. but would never demonstrate under scrutiny. he uploaded a lot of crazy clips to youtube to show his experiments but some of them he was hanging plastic models off strings and just totally faking the whole thing. his channel is still around. he wears dresses and calls himself "KARLA", and yes this story is all real.
@@Sanchuniathon384 There are cases where the compression is infinite. If you take non-repeating tiling, it's all encoded in the tile itself.
Now, the thing is that if two people "decompress" that information, they will get something slightly different, but in some cases, parts of the same tiling to one translation and rotation.
Fractals are also one good compression.
We also have the Lena compression.
With the pigeon hole principle, lossless is out of the window for 8k.
And when you consider loss, probably enough of it so that you end up with a single pixel (probably black and white) ^^
Now, with AI able to generate movies, we could achieve some compression. You would feed the compressed script, and the AI would spit out the film.
Better, an AI could compress the plot too.
I think a lot of plot would fit in 8000 characters ^^
The only mystery is why no investors listened to their engineers who knew this was impossible?
@@nix324 I'm not exactly computer literate, but from the onset it seems incredibly impractical and ufeasible, you have a computer generate pi let's say, with just two operators, and then pretty much have code or something browse through the infinite string of numbers until it finds the value its looking for, key words "looking for" it needs to know what it needs to use it, at that point wouldn't it take less space to just write down the value itself?
There's probably something in here that I'm missing but I don't understand how would this work.
Smart people always like to claim “the impossible” I believe if anything with current advances in science/ technology the more we discover the more questions arise. Be optimistic!
@@brunotonyoli9408 you dont have even the slightest clue about data compression, yet your mere "belief" without any evidence or even basic understanding of the information science behind this makes you a better judge of what is impossible or not. do you realize how egotistical this is? you think you know more than the actual people who spend their lives working on this stuff? and even though you know nothing about it?
There is very extensive statistic and stochastic research done in compression technology. Even with quality loss there a limits to how much you can compress random data. It might be possible with AI to find data streams which can be reduced to algoritms. But that's only true for a small part of all possible data streams.
If there were less words it would be easier to store them all. So u could just store a word or symbol or the most common syllables of words instead of all the individual letters? Maybe 🤷
It is definitely possible to compress all data stored on every computer in the world into 8192 bytes, just don't expect it to be useable afterwards lol.
Thats not called compression that's called hashing.
very nice take :)
he also claimed lossless
@@NavySeal2k You're a moron that thinks he's way more intelligent than remotely justifiable.
I can do it right now.... 01
Some data could have become corrupted and unreadable, sorry
I am from the Netherlands, Jan Sloot was a conman who swindled some naive and greedy businessmen into investing huge amounts of money in his too-good-to-be-true innovation.
He was the Elizabeth Holmes of computer tech. The Fifth Force, -Tesla- Theranos: selling hot air.
Thats what comes to mind straight away for me, humans realy dont think properly when you bait them with a greed carrot.
I was thinking either that or he was just batshit insane.
@@sporehux8344true... You shouldn't reject all what you didn't understand yourself but thinking that advancements make everything possible is also not right.
We see that people wich can talk in a specific way can sell everything until their ego crushes them. Damage done.
I think this is getting more common because people thinking more and more on money investments besides a bank account.
@Nordlicht05 What are you trying to say? I can't tell if you actually believe this story or are trying to agree with OP. 🤪
@@CybershamanX short....more like if you can talk, many people will give you money for the biggest shit.
Money also becomes more loose because more people try to find the next big thing and get rich.
I think the reality is that there are more frauds than geniuses. Mostly you only hear from winning story's wich makes things often worse for some.
Inside the box there was a group of very tiny filmmakers and actors, you give them a 8kb script and they spit out the movie
As a time traveller who is here on temporary business, what you describe is going to be a thing in a long future from yours but it will happen - just like nobody imagined real time raytracing would be a thing, now it is, today in your year, you have chat bots, image generators and sound generators - in a very distant future all this will be so much improved - what you have now is GARBAGE compared to the real power of AI in our time line - Real time image generation, sound, audio, real time text script, voice prompt to 3D model, real time full motion video generation, so new formats will exist which will be instructions where powerful processors will interpret and generate on the fly the representation of video and sound - so it's not really compression but real time content generation . So yeah audio and video content generation on the fly through small script files will be a thing. When doing video conferencing, the actual full video and audio data will NOT be transferred as it is in your time - in the future, the AI will sample your face and audio, and it will be re-generated live at the destination, so only minimal info is transferred, saving tons of bandwidth and solving once and for all the issue with lag and transmission errors. SO yeah, what you are seeing with tech in your year is so preliminary right now, As far as compression yes it will become far more efficient, and with much more powerful processors, compression will be able to significantly reduce file sizes, where audio and video scenes will be able to be recreated by the AI, in real time during decoding, it will be a hybrid and adaptive compression where it will be possible for parts of a video to be compressed, other parts to be converted to code which can be interpreted to recreate specific scenes and audio
Now compressing something in 8KB, technically no today in your timeline you could not compress a full traditional video with moving images and sound into 8KB. If a video was of a simple subject talking against a still background and audio - YES in the future you can compress the entire clip into kilobytes - trough AI - what it does is take a sample of the audio, uses advanced speech recognition, it learns the image to later recreate it - along with recreating the audio much like text to speech, so you are storing the representation of content into text, code or tokens representing instructions, that are later decoded on the fly. There IS actually technology that exists to do this in your present, it's not main stream yet - Starting in the year 2059 it will become mainstream.
@@jimmydandy9364 hey, whats the news when it comes to converting brain signals into "scenes" that an AI could easily transform into a point cloud and literally relieve your memories based on a set-list of memory triggers which are custom to you as a person and stimulates your brain to successfully recover a memory step by step?
@@jimmydandy9364you will have AIs for generating movies but they still won't be 8Kb in size
@@Pythoner Depends on how detailed you want. A movie script might be longer than 8kb but you can write things like «the couple goes into an argue about politics» and the AI will generate the details in the conversation.
@@jimmydandy9364 it better be the year of the linux desktop,.,.,.,
There is an inverse correlation between computer science literacy, and the belief this tech was feasible.
With enough discarded pixels, you can pretend to encode any movie into 8 kbytes. Convincing someone who watches it to give you money, well... that's a bigger hurdle.
This technology can be done. Storing a HD movie can easily be done 8 kilobytes if you know what you are doing.
@@jordanthomas2031 no. If you claim this then you have no clue what you are doing. Just none.
@@TheLouisEric I do know. Have you heard of the library of babel ? It is an infinite library that uses procedural generating algorithms. If you go to the site you can search through the infinite library. It will find what you are looking for and then it will give you an index code. With that string of 10 letters and numbers you can email it to someone and they can then use the library to find the book. By increasing the library code to use numbers and larger documents it is possible to use it to compress TERABYTES of information into an 8 bit code . A simple text message you can send to someone. That text would contain the string of numbers and letters that is its index code in the library. So you are wrong. Now what I described is just the surface of what the code can actually be used and optimized for. You do not know the truth. You are being lied to. What the military has in its classified tech sectors would blow your mind. Wake up.
@jordanthomas2031 he's right on this one
This was a hoax. If the inventor believed their own hoax, that's sad, but that just means he was a deluded individual instead of a fraudster.
Its a bit like the film A serious Man by Cohen brothers.
Ok glowie
@@hrr597You believe th claims? It's literally mathematically impossible.
Scientology founder eventually believed in his lies before dying .
@@Mikewee777good, I hope he suffered.
Having dealt with somebody paranoid like this makes me think it was all in the man's head.
I am reminded of the Starlite invention, and the magnetic counter-rotating perpetual machine invention, and how their inventors acted all squirrelly ostensibly due to paranoia of it being stolen, but actually paranoia of being found out.
Yep, TempleOS is another high profile example. TempleOS was actually pretty impressive, but there are similarities with the mental illness and paranoid delusions.
I wonder what compression he used to store it all in his head.
What if it was both? What was that movie? About the Schizophrenic Math Genius? A Beautiful Mind? John Nash. The man did groundbreaking things with in mathematics, and his contributions to game theory are incredible. Maybe this man actually did see a better way to do something and his paranoia peaked, causing him to take incredible lengths to hide his actual work, and the stress of protecting it for the last couple of days popped the ticker. Of course, it's possible he was a fraud and knew he was, and the paranoia was an act and the heart attack was brought on by having to put all the cards on the table soon. Flip a coin. For fifty cents, I'll give ya two more theories lol.
@@nerdjournal as someone with ADHD I believe he let his creative mind run away, notice how he kept starting things and not finishing them, getting stuck in the details, classic ADHD. Everyone has different amounts and deals with it differently, I think he got too ahead of himself, notice the common theme of him thinking smarter people would work it out - he hadn't fully worked it out, that's where the paranoia came from. But he was obsessed with it nonetheless.
Anyone who would believe a whole movie could be reduced to the size of a text file, is beyond help! 😂
A sixty frame short film with each frame consisting of 1 pixel would probably be pretty small, lol. Oh, and it's silent too.
a 3d game called .kkrieger was only like 95kb
@@jimnzhasmeper second?
Programaths above just gave several examples of infinite compression. Impossible for you doesn't mean impossible for everyone.
maybe with newer video generation models. you could plop in the script of the movie + a cfg and a seed and it might not be entirely out of the realm of possibility. You can already compress an image using stable defusion to
"🌀🌀 🌊 🏞💀, ⚔, 💥, 🎆, ☁, 🦖 🌍, 🌌, 📽,, 📚, 🎨, 🖌, , 💀, 🪞, 👻, 🦝, 🔳 🎨, 🖌,
Negative prompt: 💻😊😄
Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 8, Seed: 3224160573, Size: 512x768, Model hash: ec6f68ea63, Model: lyriel_v16_4x768, VAE: YOZORA.vae, Clip skip: 1, Version: 305def9, Token merging ratio hr: 0.5, Parser: A1111 parser, Score: 6.4"
and if you combine it wth "/bark `a dramatic music composition simulair to Carmina Burana -s7435 -t9m59s`"
you could have a movie that is able to be regenerated based on a relatively small text file.
I think its theoretically possible to send someone a text file and with the right tools them be able to reproduce the same movie you watched.
Because of how information works, if his algorithm allowed movies to be just 8kB, then the algorithm itself would need to be hundreds of megabytes or several gigabytes, and it would only be able to play a few movies. Essentially the algorithm itself would be the compressed movie and the "movie files" would functionally be references to movies stored in the algorithm. It's impossible to make something general-purpose like he's describing.
excellent profile pic
jupp. it is an infinite hashtable. ;)
It worked, you just needed a special companion 400 MB file to decode the 8Kb file.
Is there a working copy out there? Or information to replicate it?
@@KaranCheckerno, because this whole thing is fake
Even a 1TB companion (in todays standards) to recover ANY kind of file from 8KB would be a HUGE DEAL. Just imagine... You could just put that same 1TB content on every device out there and have a UNFANTHOMABLE amount of internal storage available due to this. Eventually, manufacturers would likely try to hardcode that 1TB content on a specific read only ASIC chip that is cheaper to produce and slap it into every device out there. It would become a standard chip to put into devices.
@@fusseldieb unfortunatly each individual movie needed its own special 400mb file.
@@KaranCheckerIt was a joke, about how one could do what he claimed, but with an algorithm hard coded for the movie, which would have the size of the movie minus the 8k or maybe more. Basically the joke is "I can make a megaphone with a squirrel tape and a megaphone"
I suspect he died from the stress of signing the deal. He would have had to disclose to his partners and there was no working tech to disclose.
The compression ratios stated are impossible, of course. Heck, in 1996, my company paid $10,000 to license a voice codec from a firm in Boston that could compress voice into 8 kilobits per second, do echo cancellation, and still sounded mostly like a voice… Telephone voice data is PCM at max 56 kbps … it’s done today regularly, cutting edge then though… but 56 to 8 isn’t as impressive as a 16 gb movie in 8 kb.
So…
I wonder if Pied Piper company in that HBO show was based on this story….
Silicon Valley was a great show!
@@dreejz it really was
From his claim that binary is obsolete, I think your mention of telephones and 56k is relevant- he wasn't talking about binary in the first place. But, his symbols per cell would have been massive, and just not practical for real-world media.
He also probably didn't build it (and I question if an ideal case could even get those data densities onto non-optical floppies), but the likelyhood of excess data corruption renders the fraud-or-not question superfluous, because even if he built it and it worked he still wouldn't have built something sufficiently reliable.
or he was suicided. if he had a solution (unknown) the elite surely wouldn't want it to be materialised. why you may ask. well, it would change the need for huge bandwidth technology which was in store ... it would have changed everything. that said just speculations as to whether he was suicided because of his invention or if he was a liar.
@@absalomdraconis"He wasn't talking about binary" is exactly like saying "he wasn't talking about numbers". It's just a numerical system, we could as easily say computers work in base 1024. The fundamental flaw in his fraudulent claim is mathematical, not engineering. A "cell" is just a different way to represent data, and all data can also be represented by numbers.
So basically a movie could be compressed into 8KB but the key to decompress it was ∞YB. Gotcha.
🤣
No the key to this is using several hash files and as you use more and more hash files in Series and parallel the data expands rapidly
@@andikatjacobdennis but then youre more likely to get a video that fits all the hashes and makes sense and isnt the intended movie
When you conssider that evvery posssible information is encoded somewhere in the number Pi, the 8kb file might just contain the starting position of your movie within Pi. 😅
China can do that lol even 1Bytes
I also have a new compression tech for sale. Currently being actively developed. It's called, Singularity. It can compress any amount of data to a ZERO-length four dimensional archive file. The compression works flawlessly already. We are still having some issues with the extractor. In fact, you can make a file of any name, i.e. RAMBO5.ISO and the exctractor will fill it up with the data. Once completed, that is.
That sounds amazing! Hope you'll achieve success with that extractor soon enough.
I don't know enough about this field to determine if this is a shitpost or not.
@@ZarHakkar It is and it is not.
black hoes, the ultimate vapor ware
lmao
Let's make a few assumptions.
1. A "feature length movie" is 90 minutes
So said movie would be 90*60 = 5400 seconds.
2. 8 kilobytes would be 65536 bits.
Now that would imply that the compression would use only 12 bits of data per second.
Gotta say that I think that there's more than 4096 potential descriptions of a 1 second scene.
If you separate in still frames a full movie there is a lot more information than storing a couple frames and calculating the frames in between by modifications of the frames fully stored.
@@julmoc37 You don't get it. Just try to imagine a compression method that just uses 12 bits per second for a full resolution movie. It just can't be done. Hell, you can't even encode audio only with that, yet alone video.
Not necessarily 90. They are often quite a bit longer and sometimes a little shorter.
@@julmoc37 - He said 12 bits per second, not per frame.
@@johncochran8497: With some specific inputs you could conceivably do it, but the set of the inputs (e.g. a single frame of a single color through the entire movies, and no sound), and matching need for a well-chosen encoding, are so extreme to themselves be demonstrations of why it shouldn't be taken seriously.
Seems like the "free energy" equivalent for data storage. Is Elizabeth Holmes his daughter?
That's also water powered!
@@dannymac653 😃
I would think that even treating it as a black box, the encoding scheme could've been shown to work, or not. Carry out an experiment in a shielded room, with an isolated power supply. Use shielded containers to prevent transmission of data to and from the black boxes. Have the inventor bring two functionally identical black boxes. Enclose one in a shielded box. Using an arbitrary video of your choosing, that the inventor hasn't had access to before, use the black box still out in the open to encode it, and write the data to a common medium such as a floppy disk. To ensure there's no trick built into the floppy disk, use one of your own computers to copy that floppy to one of your own floppies. Enclose the encoding machine, along with the floppy provided by the inventor, in a shielded container. Then take the previously enclosed black box, that will be used for decoding, out of its shielded container. Put your floppy, that supposedly has your video in the inventor's encoding scheme, into the decoding machine. Watch it play your video, or watch it fail. Requires no opening of the invention at all.
From his comment that binary was obsolete, I think anything he was experimenting with was probably in storage, and thus copying the floppy wouldn't work.
It was an attempt to make something like MIDI (or modules, trackers), but for video stream.
@@GrandePunto8V Which is analogous to what DCT already does when compressing video...
It wouldn't be necessary to do the black box thing because it's patently obvious to anyone slightly competent that this is impossible.
@@peter9477 Of course you're right, and this supposed lost invention is just bunk, and the paranoia of the alleged inventor should have been one of several red flags to anyone involved. Just an idea of how those who were considering investing in something like it (and preposterous as it is, some were seriously considering it), may test it beforehand while accommodating the inventors wish to keep its operating principles private.
I think what's misunderstood (but was claimed by Sloot) is not a WHOLE MOVIE would be condensed, but each frame would be reduced to 8K. If you read his patent, it's basically Commodore 64 character-graphics for a movie. One byte would represent a 16x16 region on a 1920x1080 frame. So you'd have to pick the 256 characters that could reconstruct the frame. You'd send the 256 16x16 characters ahead of time and from then on, a stream of 8K bytes per frame. The problem is it was only a germ of an idea and he didn't fully understand the implications. Even if you sent the 256 16x16 images (another 8K) for each frame, the image would be so crappy that it would be impossible to view or recognize as the original image.
Back when I first heard about this I tried to mathematically see if this would work, it didn't, you would need a massive massive hard drive with stored code on every device, It would work like the Library of Babel, you can find that online Its a book with every combination of words.
you wouldn't need a hard drive, but the point is that the data location would take as much or more space than the data itself.
tbh maybe we looking at this worng and it wasn't compression maybe it was a new way to render the movie and the move it self was just the audio and text to render the video but even then it doesn't work with binary, so maybe he made his own version of binary? I had the idea he made a new binary version something like dns allowing for more data to be stored in less numbers
tho if this was the case it explains why he acted the way he did given the kind of work 1 would need to do to make it
@@Kamida that doesn't make any sense nor would it be possible.
@Kamida I'm gonna explain why what you're saying is a little misguided, for you and anyone else confused by the other guys response. Big text incoming. tldr; we know binary is the best way to do it.
Any other "version of binary" is the SAME thing as what binary is, just in a different (pretty much always worse tasting) flavor. We choose binary because it is the most reduced way to convey digital information. Think about states. The week has 7 states, each being a day of the week. A day could have 24 states, for each hour it could be. An hour 60, a minute 60, a second... well you see what I'm saying. If you want to store the time, you have to write these states down somehow. We use a 10-digit, or 10-state, set of symbols because they're convenient to write and read with human hands and eyes. These kinds of symbols aren't very good for digital storage though, I'll get into that. Hopefully you follow me, I'm going to use that concept of states a whole lot.
You can have, for example, 8 possible states represented by a single symbol. Let's say we're limited between the numbers 1 thru 8 and we can't write anything else. Or you can just have 2, 1 and 2. They can also write commas, cause if i just put a space it'd be hard to read for us. If you can say something in a single 8-state symbol, could you not just use 4 (or LESS, key idea here) of the 2-state ones? If I say, "write the number 3" the 8-stater will write "3". The two-stater might seem out of luck, but they can write "2,1". This isn't how binary to 10-digit numbers works so don't misunderstand, but let's say this 2-stater doesn't know the true power of their symbols so they just say, "add up the two up, that's 3!" By god, they did it!
Now something you must realize is that computers exist in a physical world, not on pieces of paper. There must be a physical way to keep track of each state. No matter how you chose to store this, I hope you can understand that if you want to keep track of more states it will take more physical space per symbol. Imagine writing each state out. This spot is either 1,2,3,4,5,6,7, or 8. Versus: This spot is either 1 or 2. See the difference in the length of those sentences? This is unavoidable with the computers we have today and will always be until we replace them on a fundamental level. Let's forget about binary for a moment and tackle this with a clear mind.
So, if we don't want to waste space by unnecessarily accounting for more states than we always need to, how can we reduce this problem? Well... what is the least number of states we can keep track of? The bottom line, the minimal case. To save storage whenever we can. Well, it can't be 1-state, because that's no information at all. If it can't be any possibility but one thing, there's no way to make different decisions based on it. Like saying "if pigs can fly", (we assume) pigs cannot ever fly and that is the only possibility. So, you are saying that no thought is being put into it; no action will ever be done. 1 state is actually no state, its nothing.
Remember we're trying to save storage, so we want to account for as few states as possible at a time. What's the next smallest (integer) number after 1? 2. You can say "I'll get to it this afternoon". Currently it is either that afternoon or it is not. 2 states, action can be taken because you are waiting for something that actually has a chance to happen (its guaranteed in this case). You might also never do it, if you're saying something like "Only if you beat me in rock paper scissors". They can beat you or not; you do the action or not. Regardless, this "one way or the other" has meaningful information that allow decisions to be made. That quality of being able to make decisions off of something is data. Somethings holds data if that is the case. See the second definition on google and observe "the basis of reasoning
@@alkeryn1700 it's possible ur dns stores data using 4 chrs
As a person who's very familiar with how digital storage, video, audio encoding, processing and decoding work, I can tell without even watching the video:
*No.*
8 kilobytes is not even enough to store a written script of a feature film in the most compact method possible.
To test this I saved a copy of the screenplay for "2001: A Space Oddysey" and saved it to a plain text fie (not a PDF or a .DOC or .OTF - plain .TXT).
This alone takes 86 kB, so over 10 times as the claimed size of a full finished movie with sound.
With 7zip ultra compression I was able to squeeze the screenplay into 31 kB. Maybe with a custom dictionary and careful parameter tweaking using Zstandard compression you could go down to 20 kB.
And that's JUST the screenplay. The absolute best you could have in 8 kB would be a set of clues for an AI model to generate a movie for you based on. The AI model would itself probably need thousands of terabytes of data and thousands of kilowatthours to process the input, and would give you some ridiculous nonsense that'd be - at best - passable as a placeholder in the background, immediately falling apart if you give 2 seconds to look at it. And that tech would already borderline on magic.
The claim is completely ridiculous and a waste of time to even ponder.
It's like asking "can you build a city with just a hammer?".
This is of course a scam, but the comparison is more like a comression with a lookup table at the receiving end, stripped from the payload. Thus supposedly somewhat smaller. A garble of small truths probably helped to sell the scam. According to research by IBM, they achieved text compression around 90 percent.
what about KGB Archiver?
"The claim is completely ridiculous and a waste of time to even ponder"... and yet you did, even to the point of making an elaborate example to exemplify to others why it can't work. 😂
@@ButterfatFarms He saved others the trouble of pondering, taking one for the team, bless
@@makeavoyhe did the math so others don't have to.
Full props to Anomaly for pronouncing Nieuwegein perfectly as a Dutch person! I`m impressed.
This was almost as good as using carbon fibre to build a deep sea sub.
The sub was compressed to 8kb
💀💀💀
Two weeks later, still too soon . . .
An implosive comment, good Anon. That really sucked the air out of me.
@@calanon534 you crushed it with that response.
Two brief notes;
1) I had to make extensive use of a single documentary, 'De Broncode', to accrue footage for this topic. I always prefer to use multiple sources to get a fuller perspective, but in the case of visuals that was simply not possible at this time. In the future I will be more picky about topics to ensure I don't find myself in this situation halfway through production again.
2) I know I said 'Pieper' wrong. I had a tough time trying to say many of the Dutch names fluently, and ironically I screwed up one of the most simple. I was unable to go back and re-record over the holidays so I'll just put this out there now.
Other than that, I hope you enjoyed the video!
Great video as usual, don't stress about the little things. I'd never heard of this before and would hate to see worthy topics skipped over due to scarcity of footage.
As a Dutchman, aside fom Pieper I was pretty impressed by the accuracy of the Dutch pronounciations.
I agree with P A, please don’t let scarcity of footage scare you away from covering such an interesting topic, great job as always Anomaly, hope you had a nice Christmas
Hey man, there's another conspiracy in a somewhat similar vein as this one that might interest you.
Two South American friends, one of them an electrician working on TV signals or something like that, apparently discovered something incredible and began experimenting in their garage.
Strange flashes of extremely bright colourful light and explosions were regularly seen and heard by family members and other witnesses.
Some weeks later, both men were found dead on top of a hill wearing lead protective armour on their heads, also carrying strange notes, one of which seemed to be a bizarre language or scientific formula, another being a set of instructions that appeared to indicate that they were intending to meet someone or something that they had contacted.
The men displayed no apparent cause of death, and the notes that were held in a police safe inexplicably went missing.
Can we make again
It's simply impossible to reduce a full length movie to 8KB and it to be watchable no matter how cleverly it's compressed.
I agree but, whats interesting in context are flac files, they somewhat devolop a same principle, Wav files are uncompressed, Flac files are losseless, and are bit for bit as good as original files wav files, yet some how flac files , reduce the data by 70% if flac is possible what else is?
@@mikeriverajr4447 There is no guarantee it reduces it by 70%, a quick search says "up to", but how well you can actually compress a file depends solely on how similar the data is/how the data is structured that you're trying to compress.
@@mikeriverajr4447if you compress white noise for example flac won't reduce size, because its not magic
@@mikeriverajr4447 It's possible with great loss to quality to compress a movie to say 100MB and it would be recognisable although not really considered watchable, these days you wouldn't normally go much below 700MB for 720p and even that wouldn't be considered good quality, but 8KB, totally impossible.
@@iivarimokelainen well now that is even more interesting.
too much compression is really impossible, eventually the pure number of possibilities given the amount of bits something should take up is simply lower and most likely doesnt include your file
Not impossible using BitPacking. The file contains small encrypted codes, that each equals a 128 bit code...or greater that this. And no need to be unpacked before use....only a very fast computer.
@@caspice you certainly can't compress 2 bytes into 1 though
@@caspiceeither I don't know what you are saying or you are wrong. Those 128 bit codes have to be mapped onto what ever length the original data had. If the original data was 512 bit long then we lose (2^512) - (2^128) states that the original had. And I don't know what you meant by bit packing but according to what I understand it to be (and what a quick Google search confirmed) is very "optimistic".Since bit packing is just putting few smaller data types into one bigger. This has it's limitations since you cannot compress infinitly and it's probably already used in formats like mp4 etc. since storing 3/4 Bytes per color instead of 8 is kinda obvious.
@@alexki3135 The program/Video player will contain all of the 128 bit codes to decrypt the data. And sure, then it will need to be larger than normal.
@@caspice Bit packing doesn't solve the problem because a 128 bit code only has 2^128 possibilities which is 8 pixels if you only allow for 16 bits of color per pixel. In other words, even if you could bit pack to 128 bits... you could only address 8 pixels of variety. If you try to address more than 8 pixels you must lose information. You would lose most of the color depth per row of each image. It wouldn't even be recognizable.
Try it using any programming language... it simply cannot be done.
At around the 3:40 mark it's said that the data is stored in five separate algorithms that total nearly 400MB in size. Algorithms aren't really used for storage, they're logic chains (i.e. essentially lines of code), not databases. Also, if those algorithms are supposed to be storing the data then 400MB isn't anything shocking in terms of size for a movie, and if they're supposed to be code then 400MB of it would be huge, so I'm not sure what that sentence is supposed to mean. In any case, this was all a hoax supported by a man with a fragile mental state who was surely convinced himself it could work.
I agree that it seems like a pretty farfetched idea. Someone more knowledgeable about this stuff than myself wrote an interesting article, which I cited in the description. "This is not compression, this is merely clever information transmission economics. Or, if you still want to call it compression, then call it compression with shared resources. An important trait of regular compression is that it is self-contained; a blob of information can be compressed by anyone and decompressed by anyone else, without needing any external resource but the decompression algorithm. If, however, the sender and the receiver share some background resource, which in strict compression terms would be considered cheating, it is possible to transmit very short coded messages that merely point to the shared resource.
"
For more, see here: jansloot.telcomsoft.nl/Sources-1/More/CaptainCosmos/Not_Compression.htm#.X-pMi1VKjIU
I was mostly interested in the minds of everyone who signed up to this, so I apologize if I skimmed over the technical aspects too much.
@@AnomalyDocs Very interesting thank you. So the way I understand the explanation given in this article: each frame of a movie would be cut to pieces (redundancies eliminated) and stored on a flash drive connected to your screen. Each piece is accessible through a 24bit number. Then the whole movie can be reconstructed with the help of a table of references of sorts, containing a long sequence of those 24bit numbers. This table amounts to very little data in terms of size and is itself compressible. The table is read and each piece of data from the drive is then accessed and read. The idea would be to only have to transfer the reference table, but this table in no way contains the actual data of the movie, only clever references hence why he was never able to demonstrate his principle without a hard drive. This is better explained in the article really :) I'm no expert in compression but have been a software developer for a decade. Hope I helped!
@@BuckyDoodlyDoo Thank you!
@@AnomalyDocs every so often this has happened someone claims to have something like this even credible people or groups but then they end up walking it back or never showing proof
400 MB. would be more of a dictionary for decompression, not an algorithm.
I once also found a "revolutionary way" that would enable me to store movies at insanely small sizes. Then a couple hours later, I simply found that I'm terrible at combinatorics.
Crazy....he had a heart attack over the stress of closing a deal on something that was fake
The Dutch Theranos!
Probably expected a quick con and out but couldn't work out how to cleanly leave with as much cash as possible as it was all tied up.
even if that man made an electronic device that could generate the equivalent of a visual library of babel, it is impossible to even store the binary address of every frame of a movie (about 120K frames minimum) in 8KB.
Back in windows 98 days, I clicked select all photos and did send to floppy drive(shortcut) and gave it to a friend. I thought I discovered a secret sharing trick.
It’s a shame this video couldn’t;t have been compressed to 30 seconds 😢
The actual description of how his technology works is, now this might be difficult to believe, worse than current streaming technology. It's literally just sending the video uncompressed, but now a lookup table exists on the other end to add more complexity, no additional functionality, and bogging down network infrastructure.
He could have made millions until he got caught.
People didn't stream movies in the 90's.
@@Something9008oooooosh. You missed the point.
Nobody's saying that we were streaming movies in the 90's.
We were, however, compressing data.
@@Something9008 well not streaming through the internet, but tapes which had MPEG-1 video did exist and acted the same way as streaming nowadays
DVHS and Dtheater
Take the source data (video, for example), move it elsewhere, and have the "compressed code" contain the info needed to retrieve the obfuscated data. It's just shuffling stuff around while giving the illusion of compression.
I don't have any doubt. As soon as he claimed the data was lossless, it guaranteed that he was, at best, deluded. We already know the algorithm for data compression that is impossible, even in theory, to perform better than for lossless compression that works on any source data. If it was lossy, then there's some wiggle room, but for lossless you would have to disprove some pretty basic math before you could represent data with fewer bits than its fundamental Shannon entropy.
Just indexing the set of all reasonably possible movies would take far more numbers than we could count in the small "filesize" promised. Which means each possible id/key would link to many movies. He's basically saying the whole movie fits in the filename of the movie.
It was similar to the cinco midi sorting system. With a unique UMRN for each file. Very advanced. Also, this was just a marketing ploy/scammy concept and the idea was nobody was to check inside the device...like theranos. I don't think it is even possible in this day and age, to shrink video file size down to mere kilobytes, in fact the higher the resolution, the more storage space you need. Compressing video has been the only way to shrink size, at the expense of quality. Digital video formats don't behave like a compressed zip or 7z file. You can't just reassemble a complete HD movie from a mere 8 KB progenitor.
if you use the 8k as a seed for a number generator, and the numbers generated are the exact video format, then it would. but that's just impossible due to fundamental number theory. it's a fraud like perpetual motion
I understood that reference!
i think the premise was either to use stored resources on player devices to render out frames for the movies or something similar to a videogame where it renders out the movie in realtime
The close-up shots in this video are overwhelming
great video as always! thanks for putting in the work
Amazing work, as always!
'Hey I've got this magic hat'
'Can I see it?'
'.........No.'
Great stuff! I just discovered your channel today, I've watched the Sloot, Zelenoff and Scampia vids. Really interesting stuff. I knew Charlie Z beforehand, and that's the one that wound up in my feed. All your other topics are pretty obscure, which is refreshing. I thought I knew most of the anomalous subjects out there, but there are quite a few here I haven't heard of before!
I'm so glad to hear that. We pride ourselves on finding topics that don't seem overly saturated, or at least taking a new spin on ones that are. Stick around for the next video in a few weeks. Cheers.
> He did secure two patents for his project, though these don't provide enough information to replicate Jan's work …
And that, there, is a textbook example of a bad patent. The whole *point* of a patent is to reveal how something works with the promise of a time-limited monopoly on it. If a patent can't be used to replicate work, that patent should never have been issued. (Yes: I'm saying most patents, as in the overwhelming majority of them, should never have been issued.)
You deserve so many more subs! Keep working at it! Hope to see you up there with the best of em one day.
thank you!
Anomaly Documentaries you truly are the most underrated documentary maker on CZcams! Quality content for us always, a big thank you from all your loyal fans
That shot of Nieuwegein is literally one block from where I live!
It’s a small town 60k people and at the time Jan Sloot lived here 45-50k people. The weird thing is nothing was mentioned of him anywhere on the mainstream media. But the speculations of his achievement (with skepticism) were debated and explored in the hacker community.
I couldn't stop laughing when I read the claim. By the way, I have a LLM that outperforms GPT-4 that fits on a floppy disk. Trust me.
What a coincidence me too! Let's find investors to grab couple of millions!
Having read the books years back (in Dutch) and having read what was found after his death I can only conclude that Jan Sloot was a fraudster. There are so many shady parts about his story. It's no coincidence that he suffered a heart attack days before signing the most important contract for funding. It's simply impossible to store a whole film in a few kilobytes, there are simple limitations to compression that we still haven't broken to this day.
But his invention is not about compression.
It's not about compression. As I understand it has a tank full of the raw materials (sounds, alphabet, colors, whatever) a book, music, movie would need for anything. And the 8kb file is the pointers that pick out the ingredients from the tank to make the movie or whatever you want.
Like you want a cake. You have a fridge full of food stuff. And you have a thin cook book. The cook book calls for flour eggs and whatever and tge fridge is the big general tank of raw ingredients.
The tank might be 300mb but it's all you will ever need to make anything as long as you have the 8kb "plan"
Like MIDI
@@KeyboardBuster from what I’ve read the database contained smaller parts of full frames, therefor being able to recreate full frames by pasting smaller frames together. Smaller frames (for example 25x25 pixels) are much easier to manage than full resolution frames because there are much less possible combinations. If you would then find a way to not store the images itself but the equations to recreate (like we do with vector images) it would take much less space to store every possible option in a database. This would explain a couple of claims made by the inventor, “the invention uses color coding”, “it’s not about compression”, “it can recreate every movie that exist without loss of quality”. The only thing I cannot understand is how this would work for audio…
Reminds me of that kickstarter for a box that you could “charge up” with internet and take it with you so you could access internet on the go 😂
Ah, a hard drive.
Another banger! Good stuff as always!
It sounds a lot like Pied Piper. I wonder if the Silicon Valley series writers got their idea from Jan Sloot's claims and the Sloot Digital Coding System. 9:20 the name "Roel Pieper" is another interesting connection.
The actual name is "Roel Pieper", proper pronunciation of his name sounds like "Rule Peeper" (with a rolling R in his first name)...
I'd love to hear more about this story from the investors (such as Roel Pieper and Marcel Boekhoorn). If only they would share how much money they eventually lost and what their motives were to invest without factual proof.
Man - your videos are so addicting! Great job... this was eye opening - it would be so cool to see this revived, somehow, someway...
Sloot is the Loch Ness Monster of the programming world. The mystery is which one of his conned investors wised up first and decided to cut their losses.
nah, the programming world knows its a fraud. Investors though...
In all my years I've never heard of this. Fascinating video, thanks very much!
Same here! I'm quite surprised I've never heard of it before now!
After reading the description (yes, I'm oldschool like that), I immediately tried my hand at some math:
Assuming 65536 bits, 90 minute of film, that leaves you with around 12bits per second. At 25hz (PAL) that's ~0.5 bits per frame.
I never knew about this, I enjoyed it a lot. Nice to learn something new. Now down the rabbit hole I go.
There is a mathematical limit as to the maximum data that can be stored in any compressed data stream. Given that, even a TV image contains a minimum of 640x480 pixels x 4 bytes (for 16 million colors). Even if you find a way to discard unused colors and use maybe 2 bytes per pixel for 65,000 colors, you're still going to have 300kx2 bytes per pixel. Even with key frames and only writing changes, you're still going to need 400K for the first frame. Film has 24 frames/sec, TV either 25 (PAL) or 29 (ATSC). I doubt you could squeeze a film below an MP3, and that's a megabyte per minute.
I imagine in the future we could have a very low res compressed movie upscaled using AI (better than the current ones we have today) with such accuracy it would be like watching the original 4k movie, it would beat any compression technique, tradeoff would be that many things would be filled by the AI like small details, making those details different in each reproduction
Ah, just discovered your comment.. (already wrote something very similar) totally agree with you.
It is possible with BitPacking. The file contains small encrypted codes, that each equal to a 128 bit code. And no need to be unpacked before use.....only a very fast computer.
@@caspice No it is not possible. You can't pack 400 x 600 of TV type video into 10 bytes without cheating. It doesn't matter how "fast" the computer is. You simply cannot do it. I challenge you to grab 80,000 frames of 600x400 or 400x600 video. Use any compression you want. You can't compress it down unless the video is not really a movie, but just blank screen. In fact, just try doing it with 1 image from the internet. It can't be done. Recall that each pixel stores 2 bytes of color on a standard TV for 16.7 million possible color variations in the NTSC color space. It is not possible to address all the combinations of possible colors in 10 bytes for 400x600 pixels (240,000) pixels.
@@marcfruchtman9473 You forget that this is encrypted codes and it need a program to decrypt into useable code, that your video player can understand. Ex: F#127...which is an encrypted code, that you have decided to name this 128bit code. The program will then reconize this as a speciel 128 bit code and will translate it into the code, but without unpacking the code, if you do not need to do it, if it is only a video. But if it is game, you might need to do it.
This device belongs in a museum. Without any existing code one is left to wonder if all of these was just vaporware.
Great video, never heard of this before
I did electronic communication in college and worked decades in telecommunications. So I know a bit about compression. Movies streamed by CZcams and other providers use lossy compression. The video is analyzed and data that a human is not likely to notice is just thrown out. To compress an hour long movie into 8kb it would look like it was done in Minecraft. People would be big squares without features. How you would put audio in there as well, I have no idea. A standard landline phone is 54kb per second of data and that is just good enough quality to understand human speech. Hi fidelity music takes much more than 54kbs.
If a movie in 8kb were possible CZcams, Netflix, Spotify would be all over that technology. They would save billions in server storage and network connections.
Actually if a movie clip would be of a single person talking against a static or semi static background, it could be possible to shrink it to kilobytes - you are not compressing data, you are converting the content into text which is later decoded and allows it to be recreated live. This will be a thing in a distant future, in about 56 years from now it will be mainstream - as to COMPRESSION right now, now you cannot compress a traditional movie in 8KB, even if it were to look like MineCraft and all it would take far more than 8KB. In my future 32K video is a thing - it is extremely unrealistic to be using compression - file sizes would be enormous, which is why we use advanced content dynamic storage representation - which allows both compression and lossless encoding to occur on the same stream depending on the content. - Very advanced smart and dynamic AI is used at all times and continually in video in our future - with the very large display resolution it's a must.
@@jimmydandy9364 ww already have text to speech.
A story of when copium goes too far.
This would have solved the ridiculously large 100gb+ game file sizes today
I think this would be possible, but you'd have to have a crazy supercomputer, a "library of babel" algorithm built in to for all possible image frames, an image comparator, and then just a pointer for each frame of the movie into that, pointing into an array of infinite potential images.
Yes, it would be just like MIDI (or "modules", trackers) for an audio. You just need enormous database of pictures, textures and so on.
no it can't work even then. the pointer becomes too long.
The idea you describe could work but there's a limit to how low you can go because eventually the number referencing the specific "page" in the library of babel would need to be very large. There's always a tradeoff.
It’s like the starlite of the computer industry!
Great video, always happy to see more stuff from you! 👍
Starlite was real in the end with a youtuber reinventing it. Maybe the same will happen with the sloot machine, one day.
@@timothyhilditch Yeah, i’ve seen people trying to recreate starlite to varying degrees of success. Really impressed with what’s come out there, especially with such simple ingredients.
I dunno, 8kb for a lossless movie is ridiculous. I could maybe see it if we were looking much higher but currently it’s nuts, would set us forward at least 50-100 years, if not more.
Like open up a text document and see how for 8kb gets ya. It’s a good amount but not enough to call a gb of info.
Interesting story regardless.
Starlite was (probably) real. This is almost certainly a hoax
starlite had a working demonstration which was shown to experts.
This is very similar to the story of Starlite - A paranoid inventor who had millions on the table and ended up taking the secret to his grave, although that secret has more or less been figured out now.
'more or less' is a gulf when you consider what Starlite is. So no, they haven't figured it out, and aren't even close.
He didnt take the secret to his grave, his patent literally described how to make it, then in 2013 a company acquired the rights and produces it to this day. Man this is getting old.
@@Rayflower352 not quite, according to Wikipedia. 'According to a 2020 BBC Online release in the BBC Reel category, Thermashield, LLC had purchased all of Ward's notes, equipment and other related materials and is working towards a viable commercial product'
I wrote whole games in 4KB , but those are all highly handcrafted, procedural and compressed algorithms. A skill-test in engineering. But there is no way to compress a movies into a few kilobytes, unless the movie was on the machine in the first place. (the harddrive that suddenly appeared in the prototype).
He likely just put a bunch of movies onto a harddrive, and then used that short key to decrypt the selected media for replay. And fooled a bunch of naiive money centered investors.
I read a book once about the guys who wrote Elite back in the 80s. Fascinating story.
Psch, it's absolutely IMPOSSIBLE without devine power for anything that looks like any semblance of a movie to come down to anything even REMOTELY close to 8192 bytes!
Wow interesting! In my Amiga days I wondered about storing data as basically large numbers, which is what data is (just a string of 0's and 1's), and then finding some way to compress this huge number using some kind of notation. Never really figured it out, and probably isn't possible except in extreme cases.
Not too long ago I wrote a program that did something similar (and also came to the conclusion that it would be impossible to get any huge compression ratios from my method).
It worked like this...
1. Take an input of X amount of bytes.
2. Split the input into groups of 3 bytes each (2 bytes works too, 4 and above was too slow).
3. For each group, find the random seed that generates this sequence of bytes as output. (Pseudo code: rand.seed(x); if (rand()%255==a && rand()%255==b && rand()%255=c) { found(x); } else { x++; })
4. Store the seed and proceed to process the next group of bytes.
@@PhaaxGames i thought of that once! I could never get it working though. Though the math involved is the type that could be optimised by quantum computing (but like the far off stuff were not gonna see for a few decades)
@@PhaaxGames interesting approach nonetheless 😎
A long time ago I came across this crack to a program that included a video of some POV rolling down a circular corridor then opening up to a large cavern of geometric shapes and holes to other corridors on the walls and the POV camera would go around the interior of the cavern looking the various things in there.
It was an impressive little video that lasted about 1-2 minutes, but most notable to me was that the file was only 100K.
That memory always popped whenever I hear about some impossible compression algorithm.
@@GeeTrieste There have been some old games in the early personal computer days that created amazing worlds by simply exploiting pseudo-RNGs rather than trying to manually design and store them. I recall the amount of work I put as a child into trying to map the dungeon of "Telengard" (the Commodore 64 edition) and a couple decades later, reading the source code and being amazed that there wasn't actually any dungeon map stored.
This reminds me of a news report in an Australian newspaper, the AGE computer news that came out once a week. This was in the '90s. Someone was asking for startup money to develop a compression algorithm that would blow everything else out of the market. I wish I had archived that bit of print because I would love to find out what happened to the guy. Nothing ever came out of it as far as I am aware. I was discussing this with a colleague recently and we were speculating whether this guy had been captured by a large company like GD, cisco or HP to develop this or whether he had disappeared into the bowels of NSA or GCHQ. What did strike me at the time was that it looked like a scam because compression efficiency was at a very high level and what this guy was claiming was just too extreme to believe.
It might be available somewhere. There are sites with scans of newspapers going back to the 19th century. (Highly compressed of course 😅)
0:32 and someone already dies. Great video. I am glad to have discovered this channel
"The Sloot family home was _literally_ torn apart..."
Oh? How so? Did they use a track-hoe? Or just some pick-axes and crow-bars, etc.?
Dude probably managed to figure out how to convert video files into text blobs ala base64 and had a mental breakdown thinking he’d solved cold fusion.
I don't know what's more absurd - Sloot's claims or that anyone actually believed them. Not to mention that anyone was foolish enough to actually give him money.
People thought the idea of an airplane was impossible at one time yet we use them everyday.
That's some awesome Vaporware.
Commenting so CZcams knows I’m engaged by this content.
Great work as always
Much love
Thumbs up so CZcams knows I’m engaged by this content.
That's funny. I spent years working in TV technology, and know it's impossible. But my brain is frantically trying to work out how he did it! (need $ at the moment) What's more interesting to me is how he managed to interest so many investors. I guess this is why, when I worked for a tech company owned by the evil Murdoch, the CEO was a PHd computer scientist with expertise in DSP. We learnt quickly that you couldn't BS him as you generally can with other (non-technical) managers.
The currency of investors are promises. This guy had a lot of them but the problem is when he would have to actually produce something to fulfill that promise.
I mean if you could somehow reverse engineer a CRC code back into the original data without requiring massive processing power through some kind of algorithm, then it could be possible. I think it was all a hoax, but entertaining the idea, he could have been using a completely different and more analog technology concept that we just haven't considered because we're stuck in the idea of bits and bytes and 1's and 0's. Some parts of its process might not have been digital at all. If you think how a fractal looks when zooming in, you could technically use that to generate different data at different "zoom" levels, treat it like a frequency, and interpret that data as this or that. You wouldn't even really need to know the actual picture of the fractal itself, you could basically do it like how light percentage is calculated, so you'd treat the fractal shapes as light and the rest as dark, and then you'd use that to infer a simple "brightness" level. No different than if you had a 100 pixel screen and only 20 pixels were active. This combined with some other methods could lead to data storage of some type, since more methods could lead to more variations to modulate from. Or you could simply use different fractal types and count each fractal as a different level of the encoder/decoder. Think for example if you had a series of kaleidoscopes, and each one represented one bit of data, and entering that number in it changed the rotation of it to display a unique but predictable visual, then you analyzed that visual and transformed it into data, using an algorithm that say this setting equals some fixed string of data. So in order to encode, you'd simply have to be able to know exactly how much to rotate it to get that exact configuration of the kaleidoscope. And since it has a finite number of possible turns, that wouldn't be too difficult. Now multiply that by multiple different kaleidoscopes, where they are combined to increase the possible variations and outcomes. Now imagine you stack all of them on top of each other and can see all the visuals overlayed on top of each other at once, and that is also used to infer data. So for example, let's say each kaleidoscope has 100 positions, and you only need to represent the setting of each with a single number, and then you have 5 stacks on top of each other, so it's all read like a combination lock number. 5-27-6-42-8. Or even 0-27-6-42-8 if you wanted to skip one kaleidoscope altogether. Now maybe you could only really use an algorithm based on these to represent a very small amount of data, maybe just a couples of bits. But then imagine you have a million kaleidoscopes, or in the real world, something that behaves on the same principle, but can be more like a tiny dot on a hard drive, or the size of a pixel in a camera. Then you'd have your million kaleidoscopes in a very small space, able to be tuned and interpreted in a similar way to get results. It's almost like a condensed babel dictionary concept, where you don't need to store all possible combinations and reference it, but where some of those combinations already exist physically and can be combined to make other combinations in a cohesive way without having to use large amounts of data to reference them, otherwise that would defeat the purpose. Now imagine doing this all in an algorithm with an analog video signal, where a simple signal puts all the kaleidoscopes on a specific setting. That signal is then your encoder/decoder, and the other algorithm interprets the kaleidoscope positions into bits and bytes, with a tiny sensor above each kaleidoscope pixel that interprets exactly what the output signal of the kaleidoscope is without actually needing to see the finer details of the patterns/structures. It would almost in effect work like binary code, but not in binary, but rather with millions of more possibilities than just 1 or 0. We do binary because it's simple to represent reliably on digital devices and in storage. Something is either on or off, empty or full. But with this concept, it would be more like quantum computing, but even more complex. We only represent things in binary because it's easy to physically do, you either have an electric charge or the absence of one, or a bit flipped or not on a hard drive or transistor, like a light switch. But we don't actually need to represent data in that way if we have a concept that can simulate more. Instead of having a glass full or empty, we can have it half full, 1/4th full, 3/4th full, etc.....the only issue is if it's something we can measure, which is obviously difficult on small scales, but again, if we can simulate it, even using some analog method, and in a reliable and fast manner, then we can essentially create a physical encoder/decoder. We could potentially use liquid crystals as "kaleidoscopes", as they reliably change physically, and with the right method of analyzing those changes, we could infer a specific voltage level as a kaleidoscope setting. Think of how a piece of polarized tempered glass has a rainbow effect depending on the angle and distance you're looking at it. That characteristic will never change, it will always look the same from a specific position, so in a way it's physically predictable if you were to analyze it based on total light intensity reflected or something like that. But we'd need something more simple, where we'd only really need to look at it from one position, but at a different axis, i.e. looking from closer or further. Kind of like if you have a DVD where the laser would move closer or further from the disc, and the signal read from it would be different depending on how close or far you analyzed it from.
@@peoplez129The CDC is only a error correction check. It does not contain the data but only a mathematical expression of bits contained. It only says if the data has remained correctly stored not what it is.
@@eleventy-seven and only as has probably remained correct. as in if one bit has changed it should fail but some other entire different bitstream can make the same check code.
like someones gonna think that they could with fast enough computer make something that just tests what results in that crc, but that doesn't actually get you the source that made the crc originally(by accident it might).
Great documentary
Truly interesting documentary
Reminds me of the demoscene 64kb challenge from back in the day
Lets say the video storage part (ultra low size for the video file) was fraud, how about playback 16 movie simultaneously(7:03), I dont think that was possible in 1999 with a small footprint device, even if it contains uncompressed video data files. How did he demonstrate that?
Fun fact at my favourite pub there is a guy claiming to have worked out what the Sloot Digital Coding system was. Of course till today no evidence has been presented by him ;)
New idea:
Make a device that copys and encrypts a movie on it, then writes an 8kb descryption key to a memory card. Tell your customer the card has the movie in just 8kb, but you gotta buy my expensive magic box to plug said card into it.
???
Profit.
Reminds me of the story of Philip Taylor Kramer the "Drummer from Iron Butterfly" (as he's always called) who claimed to develop some crazy fractal compression method and then went missing (and his body was recovered quite a few years later at the bottom of a ravine).
I thought he was found in a burnt out vehicle.
Either way, it was obviously an assassination because he developed "disruptive technology."
*bass player from Iron Butterfly (and not during the 60s either)
It always makes you wonder do some people break into forbidden technologies then get buttoned, kinda rolls into the breakaway civilisation theory
Mentally Ill people often turn up dead too. Even a genius like Nikola Tesla was susceptible to falling into mathematical fallacies (this was before he "went crazy"); he and many others have thought they "invented" things that in fact already existed under a different description, or thought that they "discovered" some mathematical secret that was actually just a notable pattern of a known function. Such people often fall into neurotic/paranoid behavior because they usually feel they have "come upon" something that needs to be hidden from others, with their developed ideas usually centering around a "flash of insight" instead of real scientific development. This can also feed into itself when people start throwing money at that person and they're clever enough to create a demo device that impresses said people. Then those "inventors" convince themselves that it's ok their device doesn't really work, because it SHOULD work (after they quietly fix a few more things). You can see this level of delusion in the Theranos device too.
It just looks like he only flipped the sides of the naming, the 400mb program code was actually the movie data.
so if it was always the same movie demoed? it was a scam.
The narrator said the machine could play 16 movies, although I'm not sure that was proven or demonstrated. I think it might just be possible to squeeze 16x90 minute very low res, say 240, videos into 400MB.
Reading up on this there was an engineer that stated " Jan Sloot never made these extreme claims himself ".
I don't know first time hearing this story.
and they say Chuck Norris can store a whole internet onto a 1.44 MB 3.5'' inch floppy.
Good work on this documentary guys. I appreciate the neutral tone and character highlights. It was very easy to follow and quite informative.
How such obvious technological fakery can dupe so many supposedly sophisticated investors will always be a mystery to me.
Seems like what you should do is rethink how sophisticated investors really are.
the only way to explain this is that the investors never consulted with actual professionals about data compression, and just jumped on something that sounded good. this happens today all the time. there was a woman who tricked JP Morgan Chase into millions of investment money by claiming to have a company that produced an online platform with a huge user base, but she just got her college professor to code an algorythm that faked engagement and users, but because JP Morgan Chase is woke, they didnt double check because they wanted to jump on the opportunity to use this politically because it was supposedly done by a woman, and only much later did they find out that it was all a fraud. this shows that even the biggest investors or investment institutions can be blinded by the illusion of massive opportunity.
New Sub! Dryden, MI
Excellent Video! I'm gonna dig and pray you have more
Thank you!
What did they know, and when did they know it? or ... life imitates art?
Sneakers (1992) - a movie in which an all-code-breaking chip is engineered by Gunther JANek, who is killed after an answering machine containing the chip is stolen.
A movie in 8 kb. Sure. This is as realistic as the "super-algorithm that can hack every computer in the world" that we keep seeing in Hollywood TV shows. If you have even the slightest, most remote understanding of computer technology and data storage, it's plain as day that this is as possible as making a cow fly. I have no sympathy for "investors" who fall for something like this.
**Attaches a jet engine to a cow** You never said how long, nor in how many pieces they needed to be in, when landing!
The theory is to replace data with computational power. For example, a 8KB file whose mathematical hash would generate a complicated exact result. It would be small to store and easy to decode with little computational power.
The problem is encoding (creating) the seed. The computational power required is infeasible.
So much misinformation in this comment section.
@@I-ONLY-BUILD-MECHS-AND-DUSTERS I get the feeling that you call what I said "misinformation" because it is too high level for you to understand.
But I'll give you the chance to prove that you know more than an expert in cryptography. What do you think is "misinformation"?
@@I-ONLY-BUILD-MECHS-AND-DUSTERS I'll give you an example. There are many hashing algorithms in existence. But one example is SHA256.
The SHA256 hash of the letter "A" (by itself) is:
559aead08264d5795d3909718cdd05abd49572e84fe55590eef31a88a08fdffd
Imagine if that result miraculously happened to match the string you were wanting to send. If that was the case, you'd only need to transmit or store one character, "A".
The problem is choosing the right hashing algorithm with a long result and choosing the input that exactly conveys the data you are trying to send/store. There are enough random possibilities (an infinite number of possibilities because hashing algorithms can continue infinitely to make infinite strings). The problem is finding the right input that gives the correct results.
This is also the mathematical premise for why it's hard to mine a Bitcoin block, but it's easy to verify a bitcoin block.
@@featherponyHAHAHAHAHAHAHAHAHA
@@I-ONLY-BUILD-MECHS-AND-DUSTERS "a 8kb file whose mathematical hash would generate a complicated exact result"
What does this mean? What do you mean "complicated exact result"? You mean the hash? Just say it's the hash. When you hash something you get a hash number, that's it. It's just an integer lmao. What you're saying makes no sense. You're just stringing together words you heard to try and sound smart. I like how when you got called out you go "iTs tOo hIGH LeVEl 4 u"
Dude, this is literally CIS 101 stuff.
I have degrees in mathematics, information science and teach adjunct.
And to think that today, with a short text prompt to some AI GAN, you can actually generate several images and even whole movies. The tricky part is getting the output exactly where you want it, and the fact that you need humanity's combined knowledge in an accessible database for it to work...
I knew the Guy. I advised him to talk to Roel Pieter at Philips, he was at our office in Utrecht, KPN, we needed his codecs, but are not the party to deal with that. Back then, it was 1995, Pieper was trying to sell the internet in Set-top boxes for Philips, a sailing fried, we both loved Fyslan
Well at least his computer system wasn't based on tulip bulbs that would have messed things even further
It could be, we had a Dutch computer brand, named Tulip, back in the 90s Tulip Computers NV
@@Blackadder75 Mr Sloot probably had a Tulip computer...
@@Rob2 Mr who?
"he felt binary was outdated" hahaha. which microprocessors was he gonna use then?
I would assume that this technique would require a large dictionary of known keys (data sequences shared between the sender and receiver), organized in such a way that combinations of the keys would enable the algorithms to produce an infinite amount of even more complex keys.
The data transferred would be indexes of the keys or instructions of how to control the algorithms.
But still, it would probably not work like Sloot claimed for complex datasets.
I found you through SA, great stuff!
Thank you goon friend
It may simply have been a misunderstanding of the nature of the program. A script for arranging scenes which are stored on larger storage media elsewhere, like stock footage for example, could be very small and would produce high quality movies. I believe he had a genuine idea but it wasn't a compression algorithm.
Yep. His "script" would be tiny, and the stock footage indexed into would be... ooh... around 5GB for the average 90 minute movie. If you wanted to store more than one such movies you'd require approximately 5GB of lookup data per movie.
@@nickryan3417 yes, if that's what it was, it wouldn't reproduce literally any movie. It was probably fraud. On a related note, I'm wondering if there's any maths which provides a minimum ratio for data compression.
@@nineteenthly Look into Claude Shannon and information theory
I finished reading the book about this yesterday, what a story! In the book the estranger daughter and son claimed they were just looking for the will. Who knows what happened.
This world is really weird. Let's hope that we get the truth in our lifetime.
Whats the book name?
@@pravinbhorge7 De Broncode written by Eric Smit
@@ChevyPlays guy was a scammer. There’s the truth.
Looking for the will sounds reasonable. There was no “compiler” to find.
Why would anyone believe this!? Even if you assume a "feature length film" is only 1 hour long, that's less than one bit per frame!
you can reduce movie any file to 1 bit but the compression dictionary have to be multiple times larger then the original file
You'd need more than one bit unless you just want to watch Barbie and Oppenheimer over and over again.