2026-03-29

01:30 AM - Facebook
At best, AI can turn humans into CEOs which is how I use AI as I'm having AI write code and do complex Google searches to help collect a bunch of information for me. AI serves as my employees. The way I talk to AI is the same way a boss talks to his staff. A boss will give the workers a series of tasks to do. The workers are hired to figure out how to get the job done. But just like with humans, AI does not always get the job done which is a challenge.

.

On the other hand, at worse, AI can also turn humans into the fat people in that cartoon Wall-E movie. Reality usually fluctuates between the two extremes.

.

Ashton Kutcher is in a show called The Beauty where people take his drug to transform into super models. But after a number of months, their bodies start thirsting for a bunch of water. Suddenly, they explode. Oddly, that is an analogy of the giant data centers running AI which requires a lot of water or else they explode.

.

AI is a beast that takes up a lot of land and water which can in theory compete with all the farmland that they're currently buying up for double the going price. They're buying up farmland for millions of dollars as they build these data centers for AI.

.

AI is a tool that people like me are using. But at the same time, AI is also a beast that is very dangerous more so potentially than even this crazy war in Iran is.

10:40 AM
Dear Gemini
Fix script. There are problems. Study the output samples below because it seems it hung up after only a few hours. Make sure the script is working. Please make sure the lasers are resting for ten seconds every ten minutes or resting for a minute every hour. Below in the output is "COUNTERS: [?-NT:0] [*-NR:0] [/-NS:37283840] [-BS:41439232]" which you can remove because I don't need those counters. But keep the other counters. Add hardware hacks or a way to wake up the drive after ten thousand rounds because that might work. After only a few hours, I woke up and saw the Terminal spit out many rounds per second as seen below in the output, the disc drive was not on, no sounds, no lights, it looks like the drive is not active, but the drive is still showing in the Gnome Disks program. I do not know if one of the problems might be the wrong seeds because one theory would be that crazy seeds might not engage the drive if the seeds are too incompatible. Try to expand the script. Try to fix the script. Try to add things because the script needs to fix itself after ten thousand rounds. I told you for weeks that the script must fix itself but you failed to add things that can help the script itself.

01:54 PM
Dear Gemini
Script is working. Expand script. Make sure script can work on any disc regardless of if it's a new or returning disc, regardless of it's a CD, DVD, or whatever else. Script must not get stuck. Must work regardless of if 0% or 99% of the bytes were already rescued, it must be able to continue from wherever it might have left off from previous runs meaning regardless of if there are any non-tried, non-trimmed, non-scraped, bad sector bytes or not, the script must play chess in trying to continue on a quest towards 100% of the bytes rescued via the fastest routes. If possible, consider adding a file that can list top seeds that might help with any run be it with the current disc or any disc whether new or returning. The idea would be a file that can help guide the script on winning strategies, winning waves, winning seeds. The idea is that the file might log top configurations, speeds, settings, sizes, tags, variables, etc. Future runs could perhaps try the advice from the file. Of course, each disc may be different, so it would only be suggestions that the script might try. But the script still should focus on the fastest routes to 100% rescued for discs. For the scrolling bash terminal output, try to add a counter that might display which tags or flags and other things were used for each round or as often as might be displayed every so often as the output scrolls. Also, it sounds like the disc drive still makes noises like as if it is stopping and going. Yes, the sounds are not as crazy as before. The current settings might be the best considering the drive goes through Fideco into my laptop. When I told you about these issues some days ago, you suggested that one of the issues would be the -d direct related tags/flags as there might be a bottleneck or a traffic jam as data travels from the disc into the computer. So, I told you okay remove the -d related tags and what not in an attempt to minimize those issues. Perhaps, the configurations for the script is the best it can be outside of plugging the dsc drive into a tower to bypass the Fideco issue which might be one of the issues. We did talk about how I have 32 GB of RAM and 8 CPU cores, I suggested you see if any of that could help. See if there is anything else you can do to help with the flow of data from discs through Fideco into my laptop. One of the issues might be scratches on discs which might be one of the reasons it sounds like the disc lasers are like stopping and going many times again and again. Please study the following sample outputs of a DVD I am rescuing with this script via a tower SATA Lite-On DVD drive through Fideo into my Ubuntu ASUS (NVIDIA/AMD) laptop. Expand script but remember the script must be flexible for any CD or DVD as it moves from telescope to microscope modes towards rescuing 100% of the bytes the fastest, you mentioned the "Stop-and-Go Protection" which is good but the script should play chess and not all discs have many scratches which means the script should be able to adapt depending on if it's in telescope or microscope modes and depending on how many scratches or issues the script might run into which means the script should be smart enough to adapt to minimize and maximize the cluster sizes, sector sizes, block sizes, other sizes, read speeds, and all other variables in order to adapt to all kinds of situations like playing chess. You already wrote that you wrote the Stop-and-Go Protection into the scrip to "Get overwhelmed by huge requests when it hits a difficult physical area" and I agree with that as long as the script tries to be as flexible and adaptable to any disc and situation. Your idea about the "Progress-Independent Logic" is really good as long as it helps the script play chess towards rescuing 100% of all the bytes off any disc. You wrote, "Kernel Read-Ahead: The optimize_io function increases the read-ahead buffer to 4096 sectors to utilize your 32GB RAM as a shock absorber." I agree with the idea of reading ahead and with anything that can help prolong the life of the disc lasers while also trying to optimize maximum speeds for trying to rescue 100% of all the bytes off discs. I agree with your "Tag Visibility" as that can help me see what the script is doing in the output for the rounds. You wrote, "NO -d (Direct I/O): As we discussed, this prevents the traffic jams through the USB adapter." I would say the script should try -d as a last resort after 1,000 stalls during the last mile or in rare situations when there are no other options left. I agree with "Media Agnostic Detection" and with unmounting one time at the beginning. I suggest moving the "HISTORY_BUFFER" file to the home directory as I keep my laptop on for weeks and temporary files are sometimes deleted even while I might be in the middle of rescuing discs. I agree with your "OMNI-STRAT LEDGER" idea as long as it helps with any disc to optimize configurations, seeds, and everything else to help teach the chess script to become more effective and efficient at finding the fastest routes through the fastest winning waves towards 100% of the bytes off discs rescued.

09:16 PM
Expand script. But derank retries in your list of millions of seeds. I recommend no mere retrying flags/tags until after round 10,000. Make sure the script keeps a counter of all rounds, I want the round counter to continue even after restarting the script in the prompt. Like if I stop at round 221 today. If I come back to it next month, I want the script to display in the terminal round 222. I want the pool of seeds to be as big as possible, I prefer millions of seeds or the more the better, but no retry related seeds until round 10K, and seeds all seeds before 10K should be relative to what is possible. In other words, if there is only one byte left not rescued, do not select a seed that says skip a billion bytes because there is only one byte left and therefore that particular seed would not be applicable. Maybe skipping a billion bytes might allow it to land on the remaining byte to rescue it in a crazy theory but I doubt it would normally work and therefore the script should not try that before round 10K or the script should not try anything that will probably not work before round 10K. Please no retries before round 10K. When the script runs across a winning seed, you wrote, "The script maintains a floppy_strat_hall_of_fame.csv. It has a 35% chance each round to" replay it which might be similar or identical to what retries are. I would prefer that these replays/retries continue until they don't work. In general, I do not care how long or short rounds are as long as the script is trying to find the fastest route towards rescuing 100% of all the bytes, please make sure it does not get stuck in really long rounds, remember please no retries before 10K rounds outside of winning replays, and really short rounds might sometimes be ineffective in some situations if the rounds are perhaps too short to really get the job done, the script should be smart enough to find the optimal lengths for speeds, the script should try to remember the optimal settings, I would prefer to keep optimal settings wherever you see fit so that the script can pull form them when it needs to. You asked, "Would you like me to add a function that automatically generates a visual MD5 hash of the recovered image every hour to ensure data integrity during the 100% push?" I say only if that does not affect the data on the image file. The length of Pre-10K rounds depend on what is the fastest route towards 100% rescued of all bytes. Now, please fix the script. The script is still unable to adapt. I told you to fix this problem for weeks and you still fail to fix many issues. For example, sometimes a disk will get over 99% rescued. It will sometimes finish going through all the bytes so that only the bad sector bytes remain. After that, the script gets stuck in a loop where it does nothing. Sometimes, after all non-tried bytes are tried, the script fails to find a remedy to that problem. The script needs to find seeds that are applicable to having 0 non-tried bytes left. The script should not run though millions of seeds that are not applicable to the specific situations that it may currently be in at any given point as the script is running. Also, if there are 0 non-tried and 0 non-trimmed bytes left, the script needs to figure out what to do. But instead, often times, your scripts fail to. Likewise, if there are not even any non-scraped bytes left, the script must adapt to try, trim, and/or scrape depending on what might work the fastest towards getting 100% of the bytes rescued. Yur job is to make sure the script stay relevant no matter what situation it finds itself in, the script must prioritize seeds based on the situations it is in. For example, the script should not just pull trimming seeds if there are no bytes available to trim. Unless if there are bytes to trim or unless if trimming is the fastest route to winning the game. Make sure the script can adapt to anything. Update the script but remember that you have many options. Just because a disk only has bad sectors left to rescue, that does not always automatically mean that the script must only scrape or retry the bad sector bytes. Why? Because there are many options. There are many seeds that might be applicable. For example, when only bad sector (bad area) bytes remain left unrescued, you can run them through any, some, or even all of the three steps of trying, trimming, and scraping. I would not call scraping the same thing as retrying. I say no retrying before round 10K outside of winning replays and outside of scraping. The script must be flexible and adaptable to the situations they run into, it should prioritize the most relevant or most applicable seeds that have the highest probability of producing winning waves towards 100% of the bytes rescued in the least amount or time or the fastest amount of time possible, the script must continue to be a chess master as it decides when to try, trim, and scrape. If scraping as faster and more effective at rescuing bytes, do it. If trying bytes again is faster, do it. When I say you should sometimes try bytes again, I don't mean simply retrying bytes (before round 10K or excluding winning replays or sometimes scraping), I mean after all the bytes were tried, trimmed, and scraped. If all the bytes on a disk was tried, trimmed, and scrapped, then the script may want to try, trim, and scrape all remaining non-rescued bytes. Tho, it should probably not do identical seeds making the trying of the bytes again not simply a retry. I am saying there are many strategies out there. The script must figure out the most effective methods towards winning. It must learn how to win. It must remember winning strategies and always continue to adapt to all kinds of situations. Expand script. Update to include things you left out that needs to be there. Include a stall counter to be next to or combined with the silence counter. In the output, it reads: "FLAGS: [-A -S -r10 --sync]" but edit that to read "FLAGS: -A -S -r10 --sync" for example, just remove those brackets. Add back the mutation counter. Look for anything else that might be missing from earlier scripts, see if any of that should be added back into this latest script. As a chess grandmaster, one of the strategies might include trying to slow things down as much as possible during the three steps of scanning, trimming, scraping. Whether the script is running through the three steps for the first time with a disk or additional times, it may want to try reducing speeds and sizes and skipping variables as it experiments with different seeds. Yes, continue to focus on milking the winning waves, it is critical to find all shortcuts towards rescuing all 100% of the bytes of a disk. I agree with your philosophy of no dead seeds as that keeps the script relevant to the situations it finds itself in. Update the script, make sure the script is compatible with any disk in any situation. The script must work on new and returning disks regardless of if 0% of the bytes of a disk was rescued or if 99% of a disk was rescued. Make sure the script can work as fast or as slow as it needs to get to to 100% of the bytes of a disk rescued. The strategy is to find the fastest route which may include going faster or slower. Strategy may need to use telescopes or microscopes depending on the situations therein. Make sure the script can work for any disk in any situation in order to win as fast as possible. Make sure the script is fully adaptable, flexible, fixable, creative, never ending, the fastest ever, the most effective, efficient, optimal, applicable, relevant, resilient, related, connected, prepared, ready, universal, extensive, evolving, able to remember winning waves. Update the script with all of this in mind. I need the best script ever. The challenge is I have a wide variety of disks, hundreds of these disks from the 1980s to roughly 2010. Some of these disks may have various disk formats and file systems. Actually, some disks might have or might have had multiple disk formats or file systems as I may have used the disks with all kinds of computers, Windows, Apple, Linux, Unix, video game formats, MS-DOS, different versions of Windows, not just Fat12 but also Fat32, all of the different types of Windows operating systems to 2010. Perhaps any and all types of files. Some files might be deleted. I may have saved and deleted different files many times. I understand some files might be totally gone if they were totally written over. Multiple people may have used these disks. My desire was first to copy as much of the disks using ddrescue as I could before experimenting on how to rescurrect files. I would love to have the best way to at least have the best image files of the disks so I can figure out how to carve them or resurrect files from them later on. I would love a code breaker or a way to decrypt or translate out the files even if they are missing connecting data that you mentioned. You can choose to try to help me with step 1 or step 2 or both. You can decide whether or not you can improve step 1 much more as you already made a good script that runs ddrescue that I could try to continue running for hours. Step 2 would be to find ways to resurrect files off the image.

Twitter Tweets
10:19 AM

Iran is NOT Iraq.

10:54 AM
Before 2019, I used to write about how Post-2025 was gonna be a challenge as that was going to be Post-Trump.

01:15 PM
Trump was wrong about Covid/Iran, it took years for Trump to pivot off the Covid Scam, Trump is slow at pivoting.

01:32 PM
Oct 7 Hamas was funded by Israel, Tucker Carlson documentary is exposing Benjamin Netanyahu, The BiBi Files.

02:36 PM
Iran went from a strike to a few hours to a few days to a few weeks and this is the same as Covid was which also TRICKED Trump. Does Trump not realize that both Iran & Israel don't want to do long-term peace-talks?

03:14 PM
I'm America-Only and I'm against sending trillions of dollars to Israel daily all of the time and BiBi is a maniac.

03:37 PM
War is a slippery slope as more and more Americans die, it's Vietnam & Iraq all over again in Iran, Trump was scammed.

03:41 PM
Digimon is to Pokemon what Beetleborg was to the Power Rangers.

04:00 PM
Did you know the American government has over 90K entities?

---

---

---