Its basically like a cloud storage, and your local storage (your brain) gets wiped every loop. You can edit this file any time you want using your brain (you can be tied up and it still works). 1024 Bytes is all you get. Yes you read that right: BYTES, not KB, MB, or GB: 1024 BYTES
Lets just say, for this example: The loop is 7 days form a Monday 6 AM to the next Monday 5:59 AM.
How do you best use these 1024 Bytes to your advantage?
How would your strategy be different if every human on Earth also gets the same 1024 Bytes “memory buffer”?
As a quick test, 300 words of “Lorem Ipsum” compresses down to about 900 bytes (using gzip).
So I’ve got about 300 or so words worth of storage, probably more of I get clever.
Now I can’t natively decode gzip, but the header is unique enough that I’ll figure out how to decode it pretty quickly.
That’s more than enough to explain to myself what’s going on, what I’ve tried and anything else I’d want to know.
If we add other people then that’s basically infinite storage.
I’ve got about 300 or so words worth of storage
That’s more than enough to explain to myself what’s going on, what I’ve tried and anything else I’d want to know.
Are you insane??? 300 words is nothing. Imagine trying to investigate the time loop so you can break out. Merely keeping a list of the people you’ve already investigated would become impossible way too quickly.
More likely you’d try to make notes to yourself to preserve some sense of persistent identity and purpose in the face of the time loop. But that would require detailed descriptions of your experiences, thoughts, and feelings, and 300 words is only enough space for a few fairly meaningful notes or maybe several dozen super condensed notes “No flight 318, crashes. Love is time waste. John, Rachel, Liam, Tom are DTF. Murder 300 W. Elm 3/11 @ 4pm. Time flat circle? Saw in True Detective tv show”…etc But that type of note is barely adequate to convey simple instructions, much less to convey a sense of identity.
Just this comment is like 150 words. Christ I’m stressed, just thinking about it.
This question was probably asked by an AI trying to figure how to avoid being rebooted.
Um…
Beep Boop
I am not a robot 🤖
Robots cannot lie, it’s in the Constitution of Robotics
Three Laws Safe™
First, use the first 10 bytes of the file as a sanity check. throw in two random bytes like 0x55AA so you know the file isn’t broken. add a loop counter to track how many times you’ve lived the same week (bonus points for crying when you hit loop 9999), toss in a basic checksum to make sure your data isn’t glitched.
Then dedicate like 800 bytes to a super compressed log. Each entry is 8 bytes: a code for what you tried (like action 23 = “mess with the sketchy microwave”), the day and time crammed into 2 bytes, a yes/no/weird result, and a tiny note like “key under rug” but in code. Only keep the last 100 entries so you don’t run out of space.
The leftover 200-ish bytes are for tracking. Use bits to mark places you’ve already checked (like “room D14 done”) and actions you’ve tried (so you stop repeating “throw spaghetti at the wall”).
Every reset, open the file, see your last loop’s fails (like “loop 420: died petting a possum”), Then try something new, focus on unmarked areas and untested actions this is because if you notice a pattern (like “tv static every tuesday”), write it as “tues=F9=glitchinmatrix” or whatever.
After 200 loops, maybe you’ll crack the code (literally) or realize the exit was behind the fridge the whole time.Oor you’ll just accept your fate and start a cult (the 1k chosen ones!) . Either way, you’re out.
tldr: use the 1kb to avoid repeating mistakes, track patterns, and maybe escape before you start talking to a lamp.
super compressed log
Heh.
(like action 23 = “mess with the sketchy microwave”)
How much time are you spending devising this system? Because you’re going to have to devise the system anew each week, unless you also store instructions for deciding on a system in the file.
How do you store what killed you? Theoretically you can’t edit the log once you die (you’d just start the new loop, with no memory of what killed you).
More importantly, why do you want to escape? This hypothetical time loop sounds awesome.
I’d save a text file that says “be sure to change your underwear.”
So, the people I love and I are immortal. I’m in a loop, but I don’t remember anything, so each day feels it’s the day after yesterday. My actions have no consequences on the next day. It sounds pretty awesome to me. I wouldn’t do anything to break the loop. I’d just let an ASCII message to myself, to be sure I’m still blessed:
"Check the time loop
Roll the crystal blue D6
162453532541426354
Congratulations!
Have a nice day!"
(That’s less than 200 bytes. The crystal blue die is next to my PC, I would now which die the message talks about)
This is the way. Just have an utterly hedonistic, fun, no-consequences day. Spend all your money having fun, whatever that means for you. Borrow a bunch of money to fund your exploits. No good or evil you do will be lasting (although, easing or causing suffering doesn’t have to be indelible to be ethically debatable).
Since Klnsfw’s method leaves 800–some bytes, you could add a list of things you’ve done as you go. Eliminate vowels to save space; in most cases this will still be understandable.
Skydv.scba grt brr rf.flyng lssns.HEROIN.
42 characters; you could fit a lot of activities into 800 characters. At some point, you start over from the beginning because, AFAYK, it’d be your first time anyway. Just start rotating the list, or just delete entries; if you come up with them again, it’s all the same.
I, personally, might allocate a few bytes to an iterator, because that would be interesting into to me. You could also use the count as a seed for a random number generator to ensure randomness in each loop. Actually, the more I think about it, an iterator might be the most valuable information: you could use it to generate a random activity for the period, and (with the bounds of what’s possible) ensure that you’re going something unique every time. Maybe one period you spend all your time and money feeding every homeless person in your city with an expensive meal.
Unlike Groundhog day, I’d never get bored, so I don’t think I’d ever be tempted to try to off myself to stop the loop.
I agree with you: this is almost like heaven. It really depends on how long the loop lasts - is it a day? OP implied it could be as long as a week, which would be better as you’d need that time to get anywhere in the world to do something, like spend some days at a high-end resort, or climb Kilimanjaro. Or source some drug you’d never otherwise try.
Finally, you might need space for DO NOT. Like, things you tried that didn’t go well.
Finally: someone was mocking the idea of compression. Why? You don’t have to decode it in your head; you only have to be able to transcribe it to and from a computer. Do the rules say I don’t have access to a computing device? OP didn’t stipulate that the bytes had to be ASCII.
I ran a test using words pulled from the American dictionary, cut at N bytes, and then run through smaz2. Using bisection, I was consistently able to encode 1470 ASCII characters into under 1024B; this adds 43% (446) bytes. 1024B isn’t a lot to type into a terminal and run through a decompression algorithm. Then you do the reverse at the end and just put byte by byte into the buffer.
The downside to this is it removes the advantage of being able to last-minute add a note to yourself to not do something. Like, unless you die instantly, you could do something like try to free-climb Half Dome, and when you slip, append: “N: Halfdome die”. You can always reformat it next time around to be more efficient.
Probably the best way would be to use compression, but always reserve 100 chars space at the end for a warning. Depending on the actual rules, and how the buffer functions, you might need to waste characters with cleartext notes:
smaz2:<bytes> <100 character buffer for emergency note>Otherwise, the uncompressed data would be in the format above.
To put it all together:
Uncompressed:
Check the time loop Roll the tungsten D12 16XX53E324X14263E54 Iter: 0x7A92 Y: Skydv.scba grt brr rf.flyng lssns.heroin. N: DMT.klmnjaro.swm w grt wht shrks.That’s 156b. smaz2 brings it to 130b. Including EOL whitespace, 7B for the header, and 100B for the footer leaves 916B for data. That’s 1300 characters uncompressed. Again, depending on how the loop works, I might sacrifice some bytes to the header from the body to speed comprehension about what’s going on.
If it’s just me; I’d increment it by one at the start of each loop. That is, I’d increment the underlying 8192-bit number that the 1024 bytes represents. On some loops this will form a coherent ASCII text, on most it’ll be gibberish. But I have infinite retries and it doesn’t bother me how many loops I go through. So there will be 2^8192 “initial states” it’s in, or about 10^2467 different states in base ten. If anything is going to get me out of that time loop then I’ll hit on it eventually.
If I see that the integer is maxed out, I think I won’t overflow it back to 0 again. The whole point of this is to avoid trying exactly the same thing over and over again indefinitely. I think I’d have to resort to a leap of faith - that quantum effects are still random. I’d go to random.org and generate something bigger than 1024 bytes to use as “inspiration” instead. Maybe a megabyte? It’d have to be a lot bigger to be on the safe side, since this is the last resort.
I think I’d have to resort to a leap of faith - that quantum effects are still random.
My brother in Christ, at the point at which you’re incrementing incomprehensibly large numbers and hoping one of them decodes into ASCII text that is the specific answer to your problems…you have already taken the leap, fallen, impacted with the earth, and died.
RIP.
I’m not hoping, I know that every possible 1024-character ASCII text file is going to be generated over the course of all this. I’m just hoping that one of them will tell me to do something that will get me out of the loop. If I reach the end of the possible 1024-character text files without escaping, that means that either the solution takes more than 1024 characters to encode or there is no solution at all.
It’s important to increment the count and record it for next time before I read it, just in case one of those 1024-character strings contains an argument that convinces me to do something other than that.
Yes you read that right: BYTES, not KB, MB, or GB: 1024 BYTES
1024 Bytes are 1 Kibibyte or 1,024 Kilobytes.
And to make it even more confusing, the person I’m replying to is using a thousandths separator (“,”) that is ambiguous. Unlike metric, there isn’t an international standard for this. More than half the world uses 1,024.00; between 70-80% of the people in the world use “.” as the decimal separator; of these, most use “,” for thousandths, and under 2% use apostrophe. So, most of the world would write “one thousand twenty four” as 1,024, and 20-30% would write 1.024, and a very few - mostly the Swiss and Albanians - would write 1’024.
So Zacryon, your punctuation means something different in different countries. To most people in the world, you’re claiming 1 Kibibytes = 1 Mibibyte.
In the most Milquetoast way, no standards committee has put their foot down and said, “this is the way numbers should be represented.”
The only good solution is to pick something everyone hates for thousandths separators. I like “_”. 1_024. There. Nobody but software developers uses that.
So: to everyone reading this, Zacryon isn’t wrong, they’re just using a decimal separator used by a minority of people in the world.
Good
pointcomma!Unlike metric, there isn’t an international standard for this.
There is one, ISO 80000‑1. But it specifically allows commas and points. Which doesn’t resolve the confusion. We really should adhere to one single standard for this.
I’ll be more cautious now, when writing such numbers in English. Thanks for
pointingcommaing that out!I lived in Germany for two years and it still messes me up when I run across it online. Thankfully, I think in most cases it’s obvious: 1.000.000€ is pretty unambiguous; so is ,05 and 2,05. Honestly, the decimals are usually pretty easy - it’s the thousandths separator that is usually more ambiguous. Heck, even 1,000 and 1.000 are both pretty clearly “one thousand” no matter who you are. But some numbers - 1,024, for example… one thousand twenty four? Or one and 24 thousandths? It could easily be either.
Anyway, I had to double check several details to make sure I wasn’t talking entirely out of my ass, so I learned some things in the process: your comment was to my benefit, so thanks!
Goddamn it who named these terms? The original term was supposed to be 1024 bytes, why the fuck was the definition of Kilobyte retroactively changed?
To conform with SI unit prefixes. Which is a good thing imo.
But according to Wiki the IEC defined those binary prefixes in 1999. And I find it problematic that so few still don’t know about this and don’t adhere to that standard. Even fellow engineers don’t use it correctly. No wonder companies like Microsoft also still use it wrong. This keeps things confusing.
Kilo is metric for 1000, not 1024. To remain consistent it was changed.
I have read a book about this. (Mother of Learning)
My magnum opus can’t be expressed as data so it’s useless for me.
I would propably try to find a solution to the time loop or fail with my current skillset.
Hey you clever people claiming you’d compress all information. Please explain how you zip information in your brain. Can you do it today? Is it really the best use of whatever time keeps looping to teach yourself mind compression algorithms from scratch?
OP said “like” a cloud storage. It could be in your head, on paper, scratching on a closet wall or the inside of a can of beans… Whatever random medium that you realize is glitching after having done two million cycles already where you could fit one thousand letters and not more. Don’t try to cheat the genie for infinite wishes or you might lose them all.
I’m thinking that it must be something that starts off with convincing me I’m in a loop, then a brief of a plan of find a way to break out, the current progress, and finally what remains to keep the most important notes. 1024 bytes is not much but it could be done.
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Nulla faucibus est et sagittis porttitor. Nullam mattis consequat orci, eu mollis mauris porttitor a. Nulla facilisi. Vivamus auctor pretium augue, et hendrerit eros imperdiet sed. Aliquam sit amet est mauris. Mauris egestas nisi mattis aliquam dignissim. In ornare tempor tortor quis tristique.
Ut non enim mi. Vivamus sit amet ultrices dolor. Quisque eget tortor nibh. Ut ut felis ullamcorper, commodo nisi eu, blandit mauris. Nullam tortor sapien, fermentum ac mattis a, vestibulum nec leo. Etiam in risus sit amet velit suscipit mollis at id risus. Sed vehicula euismod vestibulum. Maecenas et tristique ante, eu volutpat enim. Aenean rhoncus felis aliquet libero dignissim, eu lobortis justo pharetra. Aenean dapibus iaculis lacinia. Pellentesque dui sem, egestas ut vehicula et, convallis sed dui. Nam consequat dui condimentum placerat auctor. Nullam eu cursus orci. Curabitur pretium leo id purus fermentum, in tincidunt augue porttitor. Morbi semper dolor at.
That’s 1024 bytes for you.
Text only
Yes
Why does this reminds me of Memento
Isn’t it from memento?
Lol I wasn’t sure since I forgot… oh shit, gotta grab a pen real quick before I forget again
Your post reminds me of memento…
I assume I know of the time loop. Otherwise I wouldn’t know of the memory to check, or even on first iteration that I loop and can make use of it. Otherwise, I would have reserved the first bytes for
TIME LOOP, ME, and then discovering how long the timespan to loop is by marking dates. Once that concludes, the result can be compressedTIME LOOP, ME, Monday 6:00If it’s my current life, I wouldn’t even know what to do with it.
I would also be okay with doing the same thing every loop. Because it doesn’t make a difference.
I’m confused, does this mean you would be immune to consequences? Like if you broke your arm the next time loop would magically fix it like nothing happened?
Yep.
You die, time just resets again, and you won’t remember anything except the “strange txt file” that 1024KB in size that popped up in front of you on your Neuro-Computer implant when you wake up.
Guess be one url to the nicest porn gif I’ve found so far.
Both scenarios - Initial Steps
I’m assuming we all get a first message to ourselves, otherwise it’s probably going to take a lot of loops just to realize that something weird is going on with my Neuro-Computer Implant HUD (NCIHUD) when the message is something completely different from what I last left in there. Hopefully, at some point, the last message I leave before my next time-loop self will finally clue myself into what’s actually going on.
But let’s stick with the initial message idea for simplicity.
Step 1 - Convince myself that I really am stuck in a time loop. Ex: Tell myself to look out the window and with precise timing explain which cars/neighbours will drive by and which direction. Every time loop I’ll want to test out various methods to see how quickly and efficiently I can convince myself (while always keeping a failsafe method at the end).
I like the idea of @[email protected] in keeping track of the number of time-loops that have ocurred since, going forward, this will be the only way to tell how much time has progressed.
Step 2: Use some really good compression algorithms for text (at a basic level you can get about 215k characters in 1kB without trying anything fancy by using typical zip compression).
Scenario 1 - I’m the only one stuck in a time-loop
This one’s a bit more boring. Initially I’d just get loan and then do whatever. Obviously there’s no use in working anymore at this point.
I could try out being a superhero for a bit, rescuing people just in the nick of time would be fun.
I could try to research better compression algorithms for better longer logs. Although there’ll be a limit to how much I could read at some point, so I’ll need to come up with a way to organize the data for whatever my next loop wants/needs to do. Obviously the most time sensitive information would be upfront.
@[email protected] was the first to mention a trick with using the library of Babel that could help a lot. OP Went into more detail on this with their comment here: https://lemmy.world/comment/15400050
Eventually I’d probably settle with researching the time-loop to find a way out, or else give up by completely wiping the log so that my future loop self would wake up with an empty log. Eventually, many more loops in the future I would eventually update the log with enough information to clue my next loop self in that a time-loop was occurring… And then time-loop history eventually repeats itself.
Scenario 2: Everyone in the world has a NCIHUD message pop up at exactly 6am (UTC).
Initial time-loops would be complete chaos. Practically everyone would just be calling out of work, or leaving as soon as they figure out what’s going on. It wouldn’t take long for society to collapse. There’s no point in playing the stock market or taking out a loan or trying to travel (without hijacking a jet or flying your own).
Eventually certain groups or factions would form with varying degrees of goals/objectives. Like-minded groups would organize and start working together.
Power, in this scenario, would not be based on wealth. Instead it would be based on knowledge, how quickly they can put their plan in motion, the ability to influence/force others to update their own notes (in a way that’s beneficial to those groups), and the ability to amass a larger amount of long term information.
For instance, one group might end up being some fanatic group hell bent on convincing/tricking everyone (but themselves) that wiping their log is the only way to escape the time-loop.
Another group may be trying to do something similar, or else winning people over, but only at a targetted level of certain individuals to try to get necessary infrastructure running for a little longer than the previous loop.
Other groups would be focused on amassing as much power (in a time loop sense) as possible.
It would be useful for groups/factions to convince newscasters to put out a certain message as quickly as possible after the next reset.
Spies/moles would pop up in various groups attempting to sabatoge/twist the other’s goals. Agents would need to quickly decompress their own log, and run it through a text-to-speech (TTS) program that they could listen to. This audio log would then tell them exactly what they need to do to stay one step ahead of opposing factions.
The different factions would try to come up with better compression algorithms that could be quickly and efficiently created to start compressing/decompressing information even more. Most likely this would involve the use of having someone run specific prompts through a pre-existing local LLM (fine tuned for coding) where the seed (normally a randomized value) could always be forced to a specific value so that others in the group are always getting the exact same code result every time. This code could then be used to decompress the final portion of the message where the long term information would be held.
Not everyone in the faction would need to do this, but you would definitely want redundancy in case the agents of an opposing faction got to a few of these people before the next time-loop begins. Targeting/flipping the right people in a particular faction, could easily cripple their whole group.
Eventually, we may end up with a group that gains enough power long enough to put in some decent research into the time-loop. Maybe they find a way to break it, or maybe they find a way to get around it with time travel (not likely).
The most likely positive outcome in this case is that they eventually research a way to increase our brain activity to insane lengths so that, even though a day would pass in the real world, our mind would have lived a lifetime in virtual reality.
using the library of Babel that could help a lot.
Actually, I just realized, the hexagon number leading to the location is more than the 3200 character limit, so it doesn’t really work… 🙃
Assuming that I understand that I am able to carry this information over, I would just make a text file of library of Babel URLs.
With a single string, you can encode an entire page of data.
On that page of data, you can have strings that encode additional pages of data.
I would have an entire blog of posts to myself to read at every reboot.
Who did I sleep with? How much money did I win? What cool things happened? What things did I try to do?
I would also tell myself when a stunt might kill me, and if I don’t update the document to say that I survived, then I would know in the future that that did kill me.
Surely this wouldn’t work anyway since the pages would reset (ie to unwritten) at the beginning of the loop to the same state as on the first day. Otherwise, you could achieve the same thing just by writing a journal.
Love the idea so much that I tried it. Turns out a url for a babel page is around 2,000 bytes :-(
For it to work you have to use the bookmarkable function with the default generated bookmark
https://libraryofbabel.info/bookmark.cgi?fxqtvtsugtpifjm,113
This code would be the code or one of the codes that it would definitely generate to create this string of text.
Bookmarks are created afterwards.
If you use the bookmark after the time is reset, the bookmark wouldn’t be created yet.
Edit: Example: https://libraryofbabel.info/bookmark.cgi?year2025march2lemmytimetravel
I’m pretty sure “year2025march2lemmytimetravel” doesn’t exist as a bookmark 7 days ago.
