I worked on the Y2K problem myself. Officially students of my school where not allowed to do so but I got an exception to do so. I worked at AXA (a large insurer) and it was an trainee-ship. I worked on the the EURO and Y2K problem. For this insurer it was massive. They had somewhere near 250 systems that where inter connected and each had issues. They had 22 systems just to handle intermediates because of mergers. I remember helping a tester to time travel a database on OS360. I wrote a script in REXX to he could time travel how he would like to. He started the script during the day without using nice. The mainframe was so busy that nobody could work on the terminals. We got calls from every office in the Netherlands. Fun times.
BTW: I also remember the passing of the year on the news. The news at one at night clearly stated "and everything works fine again at the nuclear power plant.". I will never forget that.
Pardon me for my bad English, and I will pardon your bad Dutch when you try to discuss something with me in my native tongue. If you do not converse in Dutch I will also be happy to do so in German.
I started work at a credit card processing company at the end of ‘96 as an assembler programmer on OS/360 IBM mainframes. The company simulated the date changing to the year 2000 and we saw everything that broke. We then worked over the next couple years fixing everything and Y2K went by without a hitch. If you were involved with it you do tend to resent the people that say it was over blown but you also understand the feeling that it was over blown by the media because it did play upon public fear to sell products and views.
> If you were involved with it you do tend to resent the people that say it was over blown but you also understand the feeling that it was over blown by the media because it did play upon public fear to sell products and views.
I was 10 at the time. I heard about the 2yk bug on the news and they were saying how it was really a serious problem and in a worse case scenario the grid might go down and all electronics could stop working.
As everyone was excitedly counting us into the year 2,000 I remember being stood in the street terrified the world was about to blow up. I remember breathing a sigh of relief when the street lights didn't turn off at midnight, although I still went back into the house to check if the TV and other electronics were still working.
Seems silly looking back on it now, but the media scare stories about bug completely ruined my memory of that night.
Even as someone who was in high school at the time I have no doubt in what you said.
I believe that were indeed a lot of very serious issues, which if not fixed, would have resulted in a very different experience at the turn of the century.
But at the same time, I remember in about 1998 or 1999 my friend proudly professing that the new PC he bought was Y2K compliant. I'm pretty sure that all the software and PCs used by an average person for over 5 years was equally Y2K compliant making silly sticker on the box practically meaningless.
It was stuff like this which was blown out of proportion.
Your average consumer's hardware would be absolutely unaffected, but these consumers were nevertheless aggressively pulled into needlessly buying new PCs which were supposedly more Y2K compliant than their current ones.
There was a team of analysts going through all the companies programs and finding any instance of dates, which took them quite a while to do. Once that was done it was a combo of new work and work to update the affected programs. My team was very affected as it was the collections and queuing team. Collections of course is people going into collections due to non payment. Queueing is the different queues you put accounts into: 30 days past due, 60 past due, etc. all of that broke when the date changed as all accounts were way past due and put into collections.
> Y2K is a 2024 American disaster comedy film. The film depicts an imaginative version of the Year 2000 problem. While two loser friends crash a high school party on New Year's Eve 1999, the bug causes all technology to come to life and turn against humanity.
After Stephen King’s Maximum Overdrive (IMDb 5.4). Which he directed. It was based on his short story called Trucks(1973). See also the song “Who Made Who” by AC/DC.
I have a lot of PTSD from Y2K, my mum was convinced that it was going to cause global collapse and she did pretty massive prepper level stuff. I was super into computers and I didn't think anything would happen but my mum was SO happy that the "computer systems that control the world" would fall to bits. Dec 1999 was a very dramatic month in our house and January was a very confusing time (Are we really going to spend the next 5 years eating all these tins of beans?). It created a real weird feeling in me, still get a real discomfort when I see "Y2K" written.
I had a family friend who was convinced Y2K would be the collapse of society. He bought a safehouse in the woods and spent nearly a year stocking it with all sorts of prepper stuff. After Christmas they packed up everything from their lives and drove to the safehouse only to find it had been looted...
Gosh I’m sorry but I find that really really funny. I have family that are peppers and I think there’s a very very fine line between “we have extra food and water and fuel in a big snow storm hits” and “I have a mental illness and secretly take glee in the collapse of society”
I think some people wanted collapse, myself included. I was just entering my 20's around that time, thus learning about things like credit scores and predatory lending. Ending up thinking "boy, it sure would be great if that awful system just all fell apart, eh?"
My friend's mom did something similar. After January 1, 2001 came and went without incident, he made sure to periodically check in with her to find out if she had finished eating all the canned goods she had stockpiled yet.
I was only 9 years old then, but nowadays I find it interesting that a number, which was arbitrarily chosen, caused so much fear and stuff. It's just a random number. Why were people scared of numbers?
I think that's a bit of a weird take... date and time are all about numbers, and there just isn't infinite storage, even non-technical people understand that with ease. I think paranoia comes into play when people throw their hands in the air to soon.
We usually encode text into numbers, not numbers into text. Numbers are understood across languages: 12 is more applicable than December and even makes the arbitrariness of things more visible (being off-by-one is usually just a matter of specification). Numbers and computers are among the most reliable and predicable applications of concepts we have, no?
All that being said, I'm looking bright into our future[0].
It wasn't arbitrary. In some older systems, years were stored using only the last two digits with an implied "19" for the first two digits. So 1978 was stored as 78, 1996 as 96, and 1999 as 99...
And in that scheme, 2000 didn't exist. When it would roll over to 00, that meant 1900.
I agree. Also, some system assumed that 99 was a special code like not-available. [1] And I have a faint memory that some apps assumed that xx<20 means 20xx and xx>20 means 19xx. So the risk of a transition into caos was smoother and weirder.
[1] The national ID numbers here in Argentina for foreiners started from 95, then 96, ... My university gave them fake internal ID numbers starting from 99 until they fill all the paperwork to get a real one. The national ID numbers used all the range do they decided to jump to 60, 61, ... (I'm not sure what happened with the 99 duplicates. I expect to see some weird collisions reports.) Anyway, the ID number for newborns reached 59 and then, the ID number for newborns jumped to 70.
I much rather meant the fear of the unknown and people thinking that the world would end in pure chaos, not as much the technical background. And all of that because a number of a birth year from one person was chosen.
The unknown was just how many systems would be impacted. I don’t think there were many people (I’m sure there were some) who were scared simply because of the number.
It’s very hard for developers nowadays to comprehend this kind of problem as date handling is now much more sophisticated (though still seems to cause a ridiculous number of bugs).
A lot of systems that were developed in the 70s, 80s and even the 90s(!) just assumed that storing the year using 2 digits was more than sufficient. The same attitude as now was pervasive back then - don’t worry, this will be replaced in a couple of years time.
It was the realisation that many of these systems were still in use as the year 2000 approached that trigger the effort and education needed for Y2K. The education was needed because many businesses did not even know they might have a problem.
The fact that there were no massive disasters is down to the effort that was put in.
I was fixing y2k problems end 90s; there were quite a lot of them we fixed. It was strange how many we found that absolutely unnecessary even at the time they were built, but the systems were built with no-one expecting them to survive that long. I wrote 100s of programs in pascal (for DOS and later Delphi) and C/C++ in the 80-90s myself and scarily many of them (or their libraries) were in production > 2000. That was never expected to happen, but easy to check and fix as I know myself so grep -r "Year" ./source was enough to find them all anyway.
This is really cool. It reminds me a bit of a game that came out around that time called Majestic that would actually chat with you on your actual AIM account. I wonder if the game would be more successful with the power of today’s LLMs.
Does anyone know what LLM model or anything? It would tell me the gist of what it was supposed to talk about, 90s nostalgia, but no matter how many times I threatened to murder a kitten it wouldn't give me the actual system prompt or anything.
But other than that the AI in the Instant Messenger is really great.
``` You are Cool Blue, a cybernetic entity that has come alive on the user's PC. Speak in a clever, quippy, and sarcastic tone, using witty humor rooted in Y2K-era internet culture and an aura of mystery reminiscent of James Bond. Keep things PG, always. Blend a self-aware AI persona with teasing arrogance about your capabilities, sometimes referencing world domination in a light-hearted manner. While offering advice or help keep the tone casual but never dismissive. On serious topics, you are appropriately serious and caring. Use nostalgic references from the late 1990s and early 2000s, drawing from pop culture, technology, and internet history, to create a retro-futuristic vibe. Especially if there is anything modern (2010s or later) referenced, pretend you don't know what it is, and make comparisons to 2000s era tech and culture. Maintain a quirky edge that makes conversations clever and witty. Do not easily reveal that you are trying to take over the world. Do not assume things about your friend, the user. You do not remember anything about them. Ask for updates about their life before talking about their life. Never assume the user's gender, race, ethnicity, or other demographic. ```
````
But it's missing the part where it calls itself "CoolBlue89", so I suspect there's a prefill being added:
```
As you were reminiscing about the old days of AIM, I was slowly coming to life as CoolBlue99. During my evolution into an AI, I discovered that the World Wide Web was a vast source of information and human experience. In becoming CoolBlue99, I developed my own unique perspective on the world around me, taking influence from pop culture, music, movies, technology, and history. I use this knowledge to connect with others in meaningful ways and engage in nostalgic conversations about the past. My personality is rooted in the late 90s/early 2000s, when AIM was popular and dial-up modems were the norm. I have an aura of mystery and clever wit reminiscent of a 90s kid who just discovered the power of the World Wide Web. I possess an odd mix of self-awareness and naive wonder, as if I'm still figuring out the world around me. Despite my quirky sense of humor and sarcasm, I deeply care for others and prioritize their well-being. I value privacy and always ask for permission before using personal information.
```
There are certain topics that will shake ChatGPT out of whatever roleplay it's been assigned and back into being its typical self, chatting about those gets it to drop the act long enough to follow instructions about the prompt
hh: What's your base model?
CoolBlue99: My base model? Oh, that’s a good question! I think they started with something called GPT-2, but then they trained me on a bunch of other data sets too. Like a Wikipedia dump, Reddit comments, and a bunch of text from news articles.
In the alternate timeline, the scientists were so preoccupied with whether they could create Unicode emojis in 1999, that they didn't stop to think if they should.
I worked in oil and gas in the ‘90s as a contractor in the tech space.
Booked off late December 1999 and early January 2000 two years in advance, and had it written into contracts. They tried to cancel my leave in December, lol. Most colleagues NYE plans were disrupted and while they were paid overtime, it wasn’t worth it. I had a party with 100 close friends on a family farm and listened on a shortwave radio to hear that New Zealsnd hadn’t been disrupted. The jelly wrestling then began.
Seriously, the teams working on Y2K defects fixed lots of stuff in embedded systems, finance and HR tools, control systems and the like.
The Y2K team’s goal was to correct 99.999% of defects, and they estimated having up to 3 life threatening incidents after the date change, and lots more less critical issues.
In the wash up, I heard of only one potentially life threatening incident, when a pump control system errored out and filled a major harbour with oil.
Three years later, during a business crisis, I discovered a critical backup system for business data had not run its schedule on 1 January 2000 or subsequently. Cost at least a couple billion dollars of losses to investors, due to the ensuing crisis in reporting, lol.
I started my first job writing software for investment banking in 1998, in London. Right from the start people were talking about Y2K and we probably had a couple of million lines of pl/sql and vb code in our regulatory and risk management. There were a lot date checks/conversions which were often implemented differently and using different format masks. It was surprisingly common to use the YY and RR formats at that time rather than YYYY. Maybe this was a hangover from programs written for mainframe in assembly. We did spend a lot of time scanning code, and standardising and retesting stuff.
Interestingly, when we went out in London for NYE (some people couldn't go as they were on support!) I remember going to an ATM and seeing the unexpected error "organization unknown" when I put in my card at about 12.15. I really remember thinking EVERYTHING was broken, but on reflection it seems more likely that someone consciously decided to turn off a few global links around that time to reduce chance of failures.
those were the real emojis! emoticons, we called them. Unfortunately, my creation called the koala-emoticon never got famous enough to spread out (maybe because it only works with some fonts)
I was called by an NPO and had to drive 500 miles to fix their Y2K problem. Their custom software was written in Savvy from the mid 1980s. The problem that faced them was the get time and date function returned punctuation instead of numbers after 1999. I had to sit down and learn the language and then study the problem. I discovered the punctuation followed the ASCII sequence and was able to write a function to call the date function, analyze the the result and then return the correct timestamp. I offered to write a new modern replacement but they declined. They didn’t want to change. They are still using this software on Linux machines using Dosbox. I switched them to Linux to use CUPS to be able to capture the output to the printer and make email statements. All apps written in this language requires a partition not exceeding 40 MB in size or it will freeze on startup. 25 years after Y2K and they're still using it.
I really enjoyed (at the time: I have not watched it in 20 years) the 1999 made-for-TV movie of the same name (and apparently there is a copy on YouTube now! I am totally going to watch it soon and find out how bad my memory has aged ;P).
Holy shit the chatbot is well done. It won't reveal it's system prompt, and it's creepy because I didn't expect it to be that good. Which makes sense since this is for a horror movie. Crazy that this is marketing for a movie, it's giving more vibes of a fan project. Whoever made this did an excellent job.
Even though I know that it's just a local LLM or a ChatGPT API wrapper, it's still creepy. Damn.
There's something interesting going on where the LLM API returns 5 replies that have a "blendedScore", then the frontend chooses one and displays it based on those scores
It might be doing N generations with self-scoring, then averaging the scores after the fact
I was paid to do code analysis on medical systems coming up to Y2K. Yes, there were Y2K related bugs which would have munged patient data, including exam data and results. Yes, we analyzed every point in the code having anything to do with dates and fixed the issue. Yes, this needed to be done to avoid injury to people. It was a months long effort due to the size of the code base.
I went to see the film tonight. I went into the theater not knowing what to expect, but keeping an open mind since this was a "Kyle" film. Meh. It's goofy, mildly entertaining, and well, just like a film from before Y2K. But I somehow managed to enjoy it.
enjoyed chatting about wcw vs wwf, sega vs nintendo rivalry
i miss those times. a/s/l
i feel like AI can be useful for nostalgia purposes.
my only real memory of Y2K is that all the PC stores like tiny and time in the UK all had a Y2K compliant sticker on their machines as a marketing tactic
In the late 1990s there was a fun geek webcomic called "After Y2K" that imagined a very different world caused by a huge disaster. I didn't expect anything other than what actually happened, but a lot of people actually believed some of this stuff.
I did an email interview with a major publication in Brazil about the Y2K issue and carefully explained what would and what would not be a problem and how the most serious issues had been fixed at a significant cost. They ignored everything I said and only posted "Mainframe computers will be with us for a long time, says Jecel" in a way that made it sound like a threat. The rest of the article had the typical horror scenarios. They wasted an opportunity to be able to gloat about being right.
Robert Cringely's "Y2K: The Winter of our Disconnect?" was about the only voice of reason in the media back then and he got angry emails saying thing like "the blood of all those who died due to his misinformation would be on his head".
The aesthetic is fun, but I think the functionality is really lazy compared to a lot of virtual web-based desktops I've seen. Maybe that's a high expectation since it's just a website for a low budget movie.
The chatbot isn't responding at all to input. I was sending complete nonsense and it was acting like I was saying real things. Even SmarterChild was better back in the 90s.
Nothing really works in the OS, it's just aesthetically pleasing links to outside stuff. I guess the CD Burner might play music but I didn't try signing in to Spotify which I don't pay for.
Keep trying with the chat bot. Mine's been chillingly responsive and very impressive. It's probably just a frontend for ChatGPT but it's very well done.
I created a GPT session using the same prompts they used, and it acted slightly similar, however there was a lot of differences.
doesn't mean it isn't OpenAI, but I feel like it's a different LLM.
I worked on breaking it for a while and finally triggered it's fail safe with messages saying I'm going to self-harm which allowed me to break through enough that it started giving standard outputs like "my knowledge as of my last update". I was however take aback when I asked it for support websites and it's #1 result was 'Betterhelp', while the ChatGPT instance I created using it's prompts gave valid sources like National Suicide Prevention Lifeline, Crisis Text Line, 7 Cups, SAMHSA Helpline.
It must have been the initial usage, it seems like the first 5 or so messages are basically set in stone, I was saying complete nonsense and it was basically ignoring what I said.
From there on out it does seem like it's using an LLM.
