What could possibly go wrong? DOGE to rapidly rebuild Social Security codebase.
- Thread starter JournalBot
- Start date
I've been on these kinds of projects. The biggest issue is you have a ton of code written by people that have long since retired and/or died. So you have to figure out what the intention was for this. You get to 80-90% pretty quickly, and then all edge cases take forever to sort out.
Upvote
124
(124
/
0)
raschumacher
Ars Praetorian
They'll blame Sleepy Joe Biden, Crooked Hillary, and That Awful Negro. Oh, and Nancy Pelosi.
Upvote
77
(78
/
-1)
Where do you even deploy a system like that? Does it run on a mainframe now? What's the bidding process going to look like, and how long will the lawsuits take when the administration decides to just hand the business to AWS?
My hope is this boondoggle takes so long it runs into the midterms and nothing goes to prod before the other side can take back over and reassert the authority as a branch of government that is currently on loan to the executive branch.
My hope is this boondoggle takes so long it runs into the midterms and nothing goes to prod before the other side can take back over and reassert the authority as a branch of government that is currently on loan to the executive branch.
Upvote
27
(27
/
0)
To some degree, some of these problems are addressable by making Internet access available in some way for those who can't normally afford it. Like public libraries.
Oh, wait - they're trying to shut that down, too.
And they're also mobilizing to eliminate low-income Internet access fees.
Upvote
56
(56
/
0)
Unfortunately, the plan is to shut down the old system as quickly as possible. And only then start working on the new replacement system - maybe.
Upvote
33
(33
/
0)
Will the new software include integration with X and a mandatory follow of Musk’s posts?
Upvote
13
(16
/
-3)
DOGE is also updating voting systems to allow dead Republicans to cast a proxy ballot for all future elections, so it's covered.
Upvote
26
(33
/
-7)
He may or may not be an idiot, but there is little doubt he is an evil cartoon villain.
Upvote
27
(30
/
-3)
I wonder whether Doge will review and publicise the no bid contact that Musk’s xAI get for rebuilding the codebase?
Upvote
20
(20
/
0)
The answer to this is yes. Republicans don't really ever stray too far from the flock.
Upvote
41
(42
/
-1)
Upvote
100
(100
/
0)
Or they'll just go with household votes. You know, where the landed (white) gentry cast votes for everybody.
Upvote
38
(38
/
0)
My money (that won't be coming from Social Security) would be on yes. They will.
I'm going to maintain that stance until I am objectively proven wrong, and old and aging people turn away from MAGA and Republicans in general in huge statistical percentages.
Upvote
29
(32
/
-3)
Upvote
97
(97
/
0)
When you identify the brilliant tech entrepreneur in this scenario, let us know.
Upvote
155
(155
/
0)
"My Social Security checks stopped and I need help"
"We're sorry, all the money in the Social Security trust fund was stolen by Joe Biden to pay for abortions for trans illegal immigrant gangsters. But Trump is here to save your- sign here and you'll get enrolled in our new plan where for just $29.99/mo we'll send you enough $TRUMP to your Xbank account that you'll be able to live like a king".
"Wow, I hate the Democrats even more now- thanks for saving me!"
Upvote
149
(149
/
0)
Has anyone here been arguing that SSA systems shouldn't be modernized? That it's dangerous to keep obsolete things running until nobody knows how to maintain?
The issue here is how they are doing it. Anyone remotely competent with building software looks at 3 months, and knows that you'd be lucky to really understand the actual knock-on effects or define an architectural plan for a system with a fraction of the complexity.
What they are doing instead - is just ripping shit out without caring about the consequences and saying we'll fix it on the tail end. You want that for a system like social security? That is insane.
This is a terrible idea.
Upvote
100
(101
/
-1)
This plan reminds me of the relative youth of the DOGE team. It sounds like something I might have thought was a good idea when I was 23 and brand new to the industry. I predict that they will try and build their replacement alongside the existing system, but because of bugs and weird corner cases, they'll never be able to turn off the old system. Instead they'll run both in parallel, indefinitely.
Replacing the Assembler code and generally working on making the code-base more portable might be a better goal. I imagine they're tied to some pretty old hardware, which is expensive to keep running.
Replacing the Assembler code and generally working on making the code-base more portable might be a better goal. I imagine they're tied to some pretty old hardware, which is expensive to keep running.
Upvote
45
(46
/
-1)
Future generations are - hopefully - going to have a field day analyzing American culture and coming to the conclusion that, within the framework of American exceptionalism, "entrepreneur" really ought to be read as "prophet".
Upvote
38
(38
/
0)
You know, I was recently bemoaning my lack of eye cancer. Thanks for taking care of it.
Upvote
14
(21
/
-7)
Customer service systems are slated for restoration in Phase VII, so in keeping with the no-testing/no-data collection disease reduction success, this project will be deemed a resounding triumph.
Hey, raise your hands if you now think Biden being called out of touch by this lot was Yet Another accusation/confession thing!
Upvote
25
(25
/
0)
I have participated in some legacy software migration projects from cobol/pl1 to cloud.
One of the approaches you can take is to automatically translate the existing code into a more modern platform. The downside is that you will end up with a codebase that has similar level of complexity as the current one. This will include boilerplate code, bespoke error handling logic and many things that weren't figured out back when the original system was created. It won't be easier to maintain either.
Modern technologies with a higher abstraction level allow you to code the same logic with fever lines of code. There's a catch though - you need to understand the logic to 'paraphrase' it in a modern manner. And often this means a lengthy reverse engineering process.
On top of that, there's qualification testing. With so many unknowns, you will most likely want to run the old and new system in parallel to compare if they behave in the same manner. Creating test scenarios, test data and automating the whole process might be even more time-consuming than the actual development.
And even if you do everything right, you might miss something, e.g. https://en.wikipedia.org/wiki/Danske_Bank#Losing_customers'_debt_records
If DOGE can pull off the migration in 6 months, it will be a true feat. I personally don't think it's likely to happen.
One of the approaches you can take is to automatically translate the existing code into a more modern platform. The downside is that you will end up with a codebase that has similar level of complexity as the current one. This will include boilerplate code, bespoke error handling logic and many things that weren't figured out back when the original system was created. It won't be easier to maintain either.
Modern technologies with a higher abstraction level allow you to code the same logic with fever lines of code. There's a catch though - you need to understand the logic to 'paraphrase' it in a modern manner. And often this means a lengthy reverse engineering process.
On top of that, there's qualification testing. With so many unknowns, you will most likely want to run the old and new system in parallel to compare if they behave in the same manner. Creating test scenarios, test data and automating the whole process might be even more time-consuming than the actual development.
And even if you do everything right, you might miss something, e.g. https://en.wikipedia.org/wiki/Danske_Bank#Losing_customers'_debt_records
If DOGE can pull off the migration in 6 months, it will be a true feat. I personally don't think it's likely to happen.
Upvote
89
(89
/
0)
Upvote
10
(14
/
-4)
My mother was a teacher in Colorado, and earned a pension out of her years of work. She also paid into Social Security, of course. But Colorado had a law on the books specifying that anyone who received a state pension was ineligible for Social Security. When she retired, she was not allowed to benefit from the system she paid into for her entire working career.
Last December, Biden signed in the Social Security Fairness Act. It superseded laws like Colorado's, rendering them null in all fourteen states that had them. My mother, and a bunch of other people, were suddenly eligible for Social Security again. Hooray!
Of course they did have to file to claim the benefit, and it took a little while for them to get the system set up for that. So recently, Mom (now 75 years old) applied for Social Security. She went to the web form and filled it out question by question, till she came to one that said: "Were you born inside of the United States?"
She clicked "Yes." Because she was.
The next question was: "Which state or territory were you born in?" And it offered a drop-down list of all the states and territories.
Friends, she was born in Washington, D. C. It's our nation's capitol, but it's not a state or territory, and therefore did not appear on the list. She was forced to either pick a state and lie or choose "None of the above." She opted for the latter, whereupon the form told her "There is a problem with your citizenship" and refused to proceed. It took several weeks of protracted phone calls to get that sorted out.
I have no evidence to suggest that this is in any way related to DOGE. The work on that form was probably done mostly before the inauguration and the subsequent creation of DOGE. But it's exactly the kind of error that I would expect to see coming from DOGE's coders. Only writ a thousand times larger, because they're aiming to replace the entire system, not just tack on a new form to handle a new class of people.
For now, Mom is scheduled to start to get her social security benefits next month. I just hope that she -- and all the other recipients -- are still getting them at the end of the year. It would be a bitter pill to finally get your social security after years of unfair denial, only to lose them again because DOGE screwed something up.
Last December, Biden signed in the Social Security Fairness Act. It superseded laws like Colorado's, rendering them null in all fourteen states that had them. My mother, and a bunch of other people, were suddenly eligible for Social Security again. Hooray!
Of course they did have to file to claim the benefit, and it took a little while for them to get the system set up for that. So recently, Mom (now 75 years old) applied for Social Security. She went to the web form and filled it out question by question, till she came to one that said: "Were you born inside of the United States?"
She clicked "Yes." Because she was.
The next question was: "Which state or territory were you born in?" And it offered a drop-down list of all the states and territories.
Friends, she was born in Washington, D. C. It's our nation's capitol, but it's not a state or territory, and therefore did not appear on the list. She was forced to either pick a state and lie or choose "None of the above." She opted for the latter, whereupon the form told her "There is a problem with your citizenship" and refused to proceed. It took several weeks of protracted phone calls to get that sorted out.
I have no evidence to suggest that this is in any way related to DOGE. The work on that form was probably done mostly before the inauguration and the subsequent creation of DOGE. But it's exactly the kind of error that I would expect to see coming from DOGE's coders. Only writ a thousand times larger, because they're aiming to replace the entire system, not just tack on a new form to handle a new class of people.
For now, Mom is scheduled to start to get her social security benefits next month. I just hope that she -- and all the other recipients -- are still getting them at the end of the year. It would be a bitter pill to finally get your social security after years of unfair denial, only to lose them again because DOGE screwed something up.
Upvote
147
(147
/
0)
No one meeting that description is currently involved with DOGE, but out of idle curiosity, why do you ask?
Upvote
71
(71
/
0)
which was written in COBOL and Assembler
If they actual start down this road, we're doomed. I can imagine that 6 weeks after it starts, Musk comes in, demands it goes live and...well all those boomers who kept saying they were never gonna see Social Security payments will get what they wanted,
If they actual start down this road, we're doomed. I can imagine that 6 weeks after it starts, Musk comes in, demands it goes live and...well all those boomers who kept saying they were never gonna see Social Security payments will get what they wanted,
Upvote
27
(27
/
0)
"...brilliant tech entrepreneur..." Musk's only talent was being such a failure at PayPal that they gave him a bunch of stock to get him out the door. Since then he's just failed upward.
"...a diverse record of extreme success..." Like real full self-driving, tunnels where cars can drive at high speeds while avoiding traffic, high speed travel in tubes that hold a vacuum, and moon landings by 2023. Loads of !!!EXTREME!!! success there.
Musk is P.T. Barnum, only without the charisma.
Upvote
75
(77
/
-2)
How long til they turn their sites on the constitution's inefficiencies? "How can a government be efficiently run off of one analog document?!?! What we need is a completely rewritten digital constitution fortified with protective AI and blockchain technology"
Upvote
19
(19
/
0)
Guessing Elon is going to have his AI do most if not all of it (hence the crazy timeline). That said, even with AI getting you 90% of the way, it sounds like an absolute disaster in the making, with ongoing effects being felt just in time for the 2026 elections.
The fact that the administration seems to have no fear of all the short to mid term chickens they've sent (and continue to send) home to roost - makes one wonder if they're counting on not having fair elections in 2026 & 2028 (otherwise they should be mortified at what they've done so far...because the path they keep choosing may get the Dems the Senate back in 2026).
The fact that the administration seems to have no fear of all the short to mid term chickens they've sent (and continue to send) home to roost - makes one wonder if they're counting on not having fair elections in 2026 & 2028 (otherwise they should be mortified at what they've done so far...because the path they keep choosing may get the Dems the Senate back in 2026).
Upvote
27
(27
/
0)
And when the melt down of the US economy takes down the rest of the world will you still be laughing?
Upvote
42
(45
/
-3)
- Add bookmark
- Featured
- #116
I worked as a program manager years ago for my employer, a Fortune 50 company, on an effort to get us off of our mainframe, with a COBOL footprint that was similar in size to this one. Let me tell you, this absolutely cannot be done in three months, AI or no AI. There are all sort of issues and complexities when you move from a centralized system to a distributed one. Think race conditions, batch vs streaming processing, totally different inter-app messaging techniques. COBOL itself is a tricky language, with "gotchas" that don't exist in other languages.
The best outcome is that they keep the old system up and running while they screw around with trying to conjure a replacement. The worst is if they start turning off pieces (or the whole thing) from the old system without testing their new systems and instead "test in prod". Knowing Trump and Musk, I don't have high hopes.
The best outcome is that they keep the old system up and running while they screw around with trying to conjure a replacement. The worst is if they start turning off pieces (or the whole thing) from the old system without testing their new systems and instead "test in prod". Knowing Trump and Musk, I don't have high hopes.
Upvote
115
(115
/
0)
The Midterms aren't gonna save us. Unless you have some math that flips the Senate.
Upvote
26
(27
/
-1)
- Add bookmark
- Featured
- #118
With experience doing enterprise platform conversions for 30 years, both from a system vendor's perspective, and as an enterprise IT architect and executive, I cannot begin to describe how guaranteed to fail, in terms of service delivery, consistency, security, and continuity, this effort is. System and language choices matter, to be sure, but they are not by themselves, or usually even primarily , determinative of success, or failure. Architecture, project management, testing, scope and quality control, transition and rollout planning ... all have a bigger impact on success or failure than any of the stuff these yahoos are going on about. You can doom a project with a bad technology choice, without question, but you can't come close to assuring it success just by making a good one.
And - I can't emphasize this enough - even if they get everything right technically, architecturally, and project-wise, it won't affect the rate of fraud or error, unless the source of fraud and error are understood, and mechanisms for detecting, and correcting them are baked onto the project requirements. COBOL is not a source of fraud, and is not inherently a source of error. Given that DOGE has identified exactly no credible fraud, and very little in the way of systemic error, that element too is doomed to fail.
This reminds me of a grossly exaggerated version of multiple spectacular system conversion failures led by big consulting at various enterprises I've been involved with over the years. The only difference is the scale (literally $Trillions across the nation at stake), and the mismatch between the hired "experts" and actual understanding of the systems they are trying to replace, are both exaggerated by a couple orders of magnitude compared with even the most massive corporate consulting boondoggles.
And - I can't emphasize this enough - even if they get everything right technically, architecturally, and project-wise, it won't affect the rate of fraud or error, unless the source of fraud and error are understood, and mechanisms for detecting, and correcting them are baked onto the project requirements. COBOL is not a source of fraud, and is not inherently a source of error. Given that DOGE has identified exactly no credible fraud, and very little in the way of systemic error, that element too is doomed to fail.
This reminds me of a grossly exaggerated version of multiple spectacular system conversion failures led by big consulting at various enterprises I've been involved with over the years. The only difference is the scale (literally $Trillions across the nation at stake), and the mismatch between the hired "experts" and actual understanding of the systems they are trying to replace, are both exaggerated by a couple orders of magnitude compared with even the most massive corporate consulting boondoggles.
Last edited:
Upvote
95
(95
/
0)
Let alone doing this with no adult governance or reasonable oversight. No guardrails, no transparency.
If they persist, the first missed payment can be deemed gross criminal negligence, as it will meet all legal criteria for that, with each and every subsequent failure a new infraction.
Upvote
42
(42
/
0)
- RFK jr is working towards assuring they won’t be around to vote anyway
Upvote
34
(35
/
-1)