The Stick of Jan Sloot (2004) | Hacker News Reader

The Stick of Jan Sloot (2004)

spronck.net

I'm reasonably familiar with this case, and there's a book about it called De Broncode by Eric Smit. I haven't read it as my Dutch is extremely limited. It's been my understanding that his invention involved analogue processing (think along the lines of an electromagnetic frequency being used to transmit a data point, rather than a single bit). The article suggests it was supposed to be able to generate all possible movies but that looks like a straw man to me. There are a finite number of movies and you could restrict this further to the top (for example) 1000 movies.

There's a patent here: https://patents.google.com/patent/NL1009908C2/en It's in English but as it's new to me I haven't read it yet either.

Jan Sloot died the day before he was due to release details of his invention to investors. The timing seems suspicious. The obvious conclusions might be either suicide in anticipation of being exposed as a charlatan, or (if you're prone to conspiracy theories) murder by someone acting on behalf of companies who had invested in digital compression, but he died of a heart attack, i.e. natural causes, though no autopsy was performed, despite relatives requesting one. I don't know the known state of his health was at the time he died.

A friend once gave me the book De Broncode, as they thought it might be inspirational for me. I must have hurt their feelings a bit as I immediately explained why it was an impossible idea.

Anybody with a cursory understanding of how technology works would have punctured that particular balloon. But in the present this guy would likely be sitting right next to one E. Holmes, M. Perry or anyone of a number of other CEOs that were 'faking it until they could make it'.

Fun read.

This Jan Sloot reminds me of a conversation I once had with my dad, an electrical engineer, where he told me how many lay people he encountered throughout his career that proposed perpetual motion machines to him.

My dad would see it as his duty to educate them on the conservation of energy, but it wasn't a pleasant experience for him to burst their bubble like that.

I don't think he ever encountered anyone who refused to be educated, though.

I've had the demo in the offices of one Hugo Krop (of Textlite fame, later jailed for fraud).

It was a complete fake and I told them that - and how I thought it worked, and for a counter demo disrupted theirs with $5 worth of electronics (VHF sweep generator). They left and never came back. I jokingly call it the first tech DD I ever did.

I worked and lived in Nieuwegein at that time (the quite small Dutch city he lived and died) and I heard many things before and after his death. The book is a nice story but it was and is clear to any computer scientist that this was just nonsense. Funnily enough Sloot managed to the then Philips ceo on board who was not a dumb guy (although his investments around the bubble actually would make you think he was and this did not help).

My though was (and still is), is that this was a clever fraud that went too far and that the stress about this fraud actually killed him. He was a nervous wreck at the end which supports that theory (although his family says this was because he was feeling paranoid and threatened).

Two years after his death I was contacted by a man who said he was a friend of Sloot and in private conversation Sloot told him the secret. As I was co-owner of an IT company in Nieuwegein and had studied math, he ended up on our doorstep. Our ceo told him to go somewhere else, but I was curious so had a few meetings with him. He showed pics of him with Sloot and explained how the system worked. He said the analog story was to throw clever guessers off (he indeed was at least paranoid enough for that this man said) and the system was, in fact, digital. It was revolutionary lossless digital compression. Of course I do not know if this guy knew anything or if Sloot ever told him anything; I will never really know. But it seems to fit the scenario I thought what had happened.

He drew this all out and wanted to pay me to implement this algorithm. I said it was basically worthless but to humor him I implemented it anyway (for free; he said when he would be rich soon, he would fix that crime). The way it worked, and I suspect so did Sloot's version indeed, is that the 'compressed' version, which indeed had 10 movies (or more) compressed in 64kb, was kind of an index for the decompression data.

So, you have 10 movies of 500mb (rather normal back then), you would end up with a memory stick with 64kb file on it which are those 10 'compressed' movies and you would have a roughly 5gb decompress.exe to uncompress them.

To the spectator it looks like you stick different sticks into your computer which have 64kb files with 10s of movies or 1000s of images, all without quality loss from their original. Which was, in my opinion, the reason no one could inspect 'his device'; it was an easy sell for the secret could be stolen that way. Of course there are very easy tricks in fat32 to show a different file size for decompress.exe which could further help the ruse.

Anyway: what he suggested is not possible and cannot be possible(kolmorogov complexity?); lossless compression of countless movies in 64kb, ergo it was nonsense. But intriguing so many people fell for it.

Sloot's 'prototype' demo apparently worked but it didn't do what he said it did. He was strongly convinced that he needed just a little bit more time to make it work for real and in the meantime he probably justified his fake demo by telling himself that once it worked all would be forgiven. It's a sad story and a good reminder that the most dangerous start-up founders for investors are the ones that believe their own bullshit.

Yeah, maybe it was not an intentional scam and maybe he believed it. The guy I spoke with who claimed to know him said that his prototype did actually work to what he said. But I was not there and it still seems an interesting story.

The prototype looked like it worked, but it didn't. It was pretty clever though and even though I don't know 100% sure how they did it I had figure out enough to make them very uncomfortable.

The setup was a small device plugged into a TV, and then you could pick any one of iirc 8 movies that they would then show a sizeable chunk of. As far as it being a scam: they believed they could make it work, but there is no way that they did not know that they were presenting a rigged demo.

Article contains a good number of sloppy inaccuracies. Here is one.

> What truly amazes me is that a man as Roel Pieper, who is a professor of Computer Science no less…

> On 1 September 1999 Pieper was appointed as a professor of Electronic Commerce, a newly created chair at the faculty of informatics and technology management of the University of Twente. Pieper ended as a professor of business administration and corporate governance at the university of Twente in 2013. [1]

Information Science != Computer Science

[2] https://en.wikipedia.org/wiki/Sloot_Digital_Coding_System

Roel Pieper is a man with a long and interesting history, some successes but also a large number of spectacular failures, and not a few borderline (or even outright) scams.

Falling for this particular scam was pretty dumb, especially for someone with his connections. Note that Jan Sloot worked at Philips before Roel Pieper and him joined forces.

Dutch "informatica" = English "computer science", although it was translated as "informatics" here.

That is interesting, and seems perfectly reasonable at first, but then please explain what "Technology Management" could possibly have to do with Computer Science. Informatics (domestic meaning) fits with Technology Management, they're complementary. Computer Science, which is a subset of Mathematics, has nothing to do with management of technology (or computers). Techs, administrators and SysOps are not computer scientists, and even if they happen to be (which is becoming more common), they're not doing any computer science when they're managing technology. What is the translation of (western) informatics? Is it possible the translation is correct? [1], [2], [3], [4] Or is it more likely Dutch universities don't know what Computer Science is? Is there also a Dutch department of Veterinary Science and Pet Management somewhere? Or a department of Astronomy and Optometry? Mechanical Engineering and Machining?

[1] https://da.wikipedia.org/wiki/Roel_Pieper ??

[2] https://translate.google.com/?sl=auto&tl=nl&text=informatics...

[3] https://translate.google.com/?sl=auto&tl=nl&text=computer%20...

[4] I suspect, not that OP is necessarily incorrect, but that this is more interesting than I thought

It's become a bit of a joke at this point, the way Dutch universities give their programs and departments English titles to appear hip and international.. It seems there are (or were) 2 faculties, "Informatica", and "Technologie & Management", and the chair "Electronic Commerce" was part of both.

This is like the idea of using Pi for compression - if Pi contains all strings of digits, then we find whatever data we want to encode in Pi and store the index where it starts. But turns out the number of digits in the index is going to be (on average) greater than the number of digits pointed to in Pi anyway

I didn't see it explcitly mentioned: arithmetic coding was one of the more mind bending things to learn about. https://en.m.wikipedia.org/wiki/Arithmetic_coding

This story reminds me a bit of the modern cryptocurrency stuff - I wonder how many of those involved are like Pieper and likely know it is bullshit but continue with the charade nonetheless as a vehicle to enrich themselves at the expense of the less savvy.

The difference with crypto is that Sloot's scheme is information-theoretically impossible. For crypto the math actually is correct; whether it is viable is a societal/human matter.

Other investors back in the day passed on this and told him they believed it to be a scam, but that did not stop Pieper from taking the CEO role in the company that was to commercialize Jan Sloot's 'invention'. Pieper did a lot of stuff that was questionable (Ring!Rosa for instance), but the list is much longer.

The biggest thing he messed up was probably killing Philips as a viable brand (but to be honest that giant was already on the way down for a while).

Define bullshit.

Well, if you're a bit too bullish about your shit, your shit is probably bullshit

> Evidently, Jan Sloot was a crank. He showed all the signs attributed to cranks. He was paranoid. He felt himself an expert in a field he had no higher education in. He worked alone. He spent decades on one problem any mathematician could have told him was impossible to solve in the first place. He got angry at people who questioned him. He felt misunderstood. He had delusions of grandeur.

This describes very much some of the antivaxers I know :)

Reply on news.ycombinator.com