mtutiger Posted February 18 Posted February 18 1 hour ago, Tiger337 said: I was mostly just sharing my experience working with this kind of data There's something to be said for having experience in something and not just just machine-gunning "Dunning Kruger" takes Quote
romad1 Posted February 19 Author Posted February 19 (edited) was just reading something on the freep about Ford pushing F-150 upgrades to 2028. Now, I just skimmed the headline. I wonder at what point do the companies who have worked their asses off to sustain their place in the marketplace in the face of well, all the reasons both internal and external, just throw their hands up in the air in the face of Elon Musk deciding winners and losers without having divested himself of his deep financial interests. Edited February 19 by romad1 Quote
Tigeraholic1 Posted February 19 Posted February 19 32 minutes ago, romad1 said: was just reading something on the freep about Ford pushing F-150 upgrades to 2028. Now, I just skimmed the headline. I wonder at what point do the companies who have worked their asses off to sustain their place in the marketplace in the face of well, all the reasons both internal and external, just throw their hands up in the air in the face of Elon Musk deciding winners and losers without having divested himself of his deep financial interests. F series is the #1 selling vehicle in America. They also have a ton of next gen issues due to EPA standards. The little birdies I know at Ford are extremely happy the efficiency regulations are being stripped. Quote
Screwball Posted February 19 Posted February 19 5 hours ago, Tiger337 said: I doubt it's as high as 12 million people, but this is something I have seen in my work too. I wouldn't assume it's fraud. It could be in some cases, but it is more likely to be erroneous birthdates due to two-digit years. What I mean is if someone has a two-digit birthyear "00", they were probably born in 2000, but might be assumed to have been born in 1900 due to an old algorithm. 0000 or 9999 could also indicate a missing value. https://www.ntd.com/database-lists-12-million-people-older-than-120-eligible-for-social-security-musk_1048177.html I remember this. There were some freaked out people. Year 2000 problem Could be in the history thread. Quote
Tiger337 Posted February 19 Posted February 19 2 minutes ago, Screwball said: I remember this. There were some freaked out people. Year 2000 problem Could be in the history thread. It's still a problem in 2025. I still receive a lot of data sets with two-year dates in them and they can create a mess when birthdates and ages are involved. Quote
romad1 Posted February 19 Author Posted February 19 Insider information...some people claim that's irrefutable. Quote
Screwball Posted February 19 Posted February 19 6 minutes ago, Tiger337 said: It's still a problem in 2025. I still receive a lot of data sets with two-year dates in them and they can create a mess when birthdates and ages are involved. I'm not a programmer, but it seems to me this kind of stuff, let's call it legacy database type data, someone along the line would have found a way to update the data form so it would no longer be a problem. Quote
pfife Posted February 19 Posted February 19 (edited) The SSA is the authority on whos dead in data. Ive worked on dozens of products that check the SSA Death Master File (DMF). Elons jabronis are idiots Edited February 19 by pfife Quote
Tiger337 Posted February 19 Posted February 19 11 minutes ago, Screwball said: I'm not a programmer, but it seems to me this kind of stuff, let's call it legacy database type data, someone along the line would have found a way to update the data form so it would no longer be a problem. You would think. Any data entry form containing year should always be 4 digits. Once it gets to the programmer, there is generally no way of knowing whether a two-digit year is in the 1900s or 2000's. The problem is when you have a massive database, the data are being entered at different places using different systems. I used to manage a database with substance abuse treatment admissions from about 20 different states. All states had the same variables, but each state had it's own way of formatting variables. Then when you finally get them on the same page, one of the states decides to do it differently the next year. Quote
Screwball Posted February 19 Posted February 19 2 minutes ago, Tiger337 said: You would think. Any data entry form containing year should always be 4 digits. Once it gets to the programmer, there is generally no way of knowing whether a two-digit year is in the 1900s or 2000's. The problem is when you have a massive database, the data are being entered at different places using different systems. I used to manage a database with substance abuse treatment admissions from about 20 different states. All states had the same variables, but each state had it's own way of formatting variables. Then when you finally get them on the same page, one of the states decides to do it differently the next year. It would no doubt be a large project, but convert databases by custom code, then consolidate into the new system. We've had time. Or maybe AI will solve it... Quote
Tiger337 Posted February 19 Posted February 19 (edited) 10 minutes ago, Screwball said: It would no doubt be a large project, but convert databases by custom code, then consolidate into the new system. We've had time. Or maybe AI will solve it... AI can probably solve it, but I am not so optimistic about AI. AI will only be as good as the people who design the algorithms. Ultimately, I think AI will be designed mostly for profit just like everything else. Whatever the wealthy want is the AI world we are going to get. Somtimes that will be a good thing, other times it won't be. Edited February 19 by Tiger337 1 Quote
pfife Posted February 19 Posted February 19 The 2 digit year isnt a problem in the SSA. SSA DMF is the gold standard for death data. 1 Quote
Tiger337 Posted February 19 Posted February 19 2 minutes ago, pfife said: The 2 digit year isnt a problem in the SSA. SSA DMF is the gold standard for death data. That is good to know. I have not used that system, but I have seen the problem elsewhere especially in state systems. Quote
Screwball Posted February 19 Posted February 19 7 minutes ago, Tiger337 said: AI can probably solve it, but I am not so optimistic about AI. AI will only be as good as the people who design the algorithms. Ultimately, I think AI will be designed mostly for profit just like everything else. Whatever the wealthy want is the AI world we are going to get. Somtimes that will be a good thing, other times it won't be. I'm a lifelong computer guy, worked in IT (and related), and feel very fortunate to ride that train over the years. Put food on the table and a roof over my head. I'm also impressed with the gains in technology. Anyone remember MS-DOS? Look where we are now. Impressive stuff. But this AI stuff scares the **** out of me. 1 Quote
pfife Posted February 19 Posted February 19 1 minute ago, Tiger337 said: That is good to know. I have not used that system, but I have seen the problem elsewhere especially in state systems. Oh yeah, same. I matched 26m rows to DMF a few weeks ago using dob and its legit. Quote
Tiger337 Posted February 19 Posted February 19 Just now, Screwball said: I'm a lifelong computer guy, worked in IT (and related), and feel very fortunate to ride that train over the years. Put food on the table and a roof over my head. I'm also impressed with the gains in technology. Anyone remember MS-DOS? Look where we are now. Impressive stuff. But this AI stuff scares the **** out of me. Right, I go l go all he way back to punch cards! The gains have been amazing. The one thing I have found in recent years though is data security (which I realize is an absolute necessity) is now preventing me from getting the most out of computers. It can be paralyzing at times and it will only get worse. Quote
pfife Posted February 19 Posted February 19 I still use dos a lot. Its really fast for a lot of stuff. Quote
Tiger337 Posted February 19 Posted February 19 4 minutes ago, pfife said: I still use dos a lot. Its really fast for a lot of stuff. I use it occasionally. There is a data set of all the at bats in baseball history which I used to use a lot and the only way to update it was with DOS. I don't know if it still requires DOS. Quote
Screwball Posted February 19 Posted February 19 1 minute ago, pfife said: I still use dos a lot. Its really fast for a lot of stuff. It is. Simple example. I create a directory for each of my students so I can store their assignments. If you open file manager (old name, explorer now?), hold down the shift key and right click on the directory, you can open a "command shell" window. Looks just like the original DOS window. Type MKDIR, put in a bunch of names separated by commas and you instantly have that many directories. Easy button defined. Funny, I can't do this at school. The IT nazi's have that little trick restricted, but have an ocean of other stuff dicked up to the highest degree. Go figger... Quote
pfife Posted February 19 Posted February 19 3 minutes ago, Screwball said: It is. Simple example. I create a directory for each of my students so I can store their assignments. If you open file manager (old name, explorer now?), hold down the shift key and right click on the directory, you can open a "command shell" window. Looks just like the original DOS window. Type MKDIR, put in a bunch of names separated by commas and you instantly have that many directories. Easy button defined. Funny, I can't do this at school. The IT nazi's have that little trick restricted, but have an ocean of other stuff dicked up to the highest degree. Go figger... I think you can go to the dir in file explorer then type cmd in the address bar and it will open a cmd prompt in that dir They might not have that way blocked! Quote
Screwball Posted February 19 Posted February 19 10 minutes ago, pfife said: I think you can go to the dir in file explorer then type cmd in the address bar and it will open a cmd prompt in that dir They might not have that way blocked! SWEET!!!! Great tip! It worked here at home, let's see how the IT nazi's stack up. Thank you. 1 Quote
gehringer_2 Posted February 19 Posted February 19 (edited) 7 hours ago, Tiger337 said: I was mostly just sharing my experience working with this kind of data exactly - sometimes the data doesn't imply what you think it does. Fro example, there could be various reasons book keeping on some things aren't up to date, doesn't necessarily mean any money is going out because of it. Edited February 19 by gehringer_2 Quote
gehringer_2 Posted February 19 Posted February 19 1 hour ago, Screwball said: SWEET!!!! Great tip! It worked here at home, let's see how the IT nazi's stack up. Thank you. you can even use a bunch of LInux CLI syntax now in Windows Powershell - thing is I never know what isn't going to work. Quote
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.