40 comments

[ 0.32 ms ] story [ 52.8 ms ] thread
Surprised IBM and Oracle have not started on their own frontier models. Or maybe they have?
"After all, if Dario Amodei had bought puts on IBM, and the dozens of companies that have plunged more than double digits in recent weeks, he would have made billions, certainly enough to fund his company for months if not years. "
Banks remain with COBOL because it's unsexy and stable. And then they say... let's just YOLO some vibe code into the next release sight unseen! Logic checks out.
Banks are slowly moving away from their old COBOL systems. It's about cost as much as it's about catching up with the neo-bank competition.

The main thing that makes this difficult is that in most cases the new system is supposed to be more capable. Transactional batch processing systems are replaced with event-based distributed systems. Much more difficult to get right.

You couldn't score any higher on the risk factors. The training corpus for COBOL can't be all that large so the models won't understand it that well. Humans are largely out of the loop and the tooling guardrails are insufficient. Causing a billion dollar disaster with the help of a "shotgun surgeon"? Priceless.
(comment deleted)
Relevant to Colombia's payment infra fragility: Bancolombia outage blocks transfers to Nequi/other banks since Feb 22 (IBM machine failed in maintenance). ~70% of national txns (100M+ /6mo, 600K interbank/mo). Daily USD flows: $50M+ A la Mano + Nequi peaks $100M+. Single-vendor risk in billion-scale retail payments? Details: https://www.bloomberglinea.com/latinoamerica/colombia/caidas...
Step 1. sell your shares Step 2. freak out retail investors Step 3. buy the dip Step 4. back to Step 1.

this has been going on all Feb.

People should also keep in mind that Anthropic got a bunch of fundraising from quant/hedge firms...
COBOL is the perfect language for LLMs because it looks just like the English text they were trained on to begin with.
LOL, anyone who thinks an LLM is smart enough to untangle 50+ years of cobol spaghetti has obviously never worked at a bank or insurance or railroad or.....
Jesus Christ the comments on that Zero Hedge blog
It's not even "Our product can write COBOL", it's "Our product can analyze your COBOL codebase and generate a plan for migrating to a new tech stack".
Finally!!

I’m porting my whole codebase to cobol!

I write SAAS suites for archeological sites.

I think I've seen 2 initiatives to move off of AS/400 to a something else in my lifetime and neither one completed. One was at a bank another at an insurance company. Not to mention that a typical COBOL programmer is more interested in retiring than learning to vibe code. At this point I think the software stocks have reached peak panic and hysteria. There is just no rhyme or reason for sharp declines like this.
So curiously I wonder if it's not that Anthropic/Claude can do this magically. More like can individuals at IBM who are heavy hitters just leave and create their own company and effectively provide these services because AI gives them the productivity to do so?
Maybe.

And then win the contracts to do this and have sufficient bankroll that they can be successfully sued and recover damages if they screw up?

No.

Someone like accenture might eat their lunch though

This seems to make the classic mistake that everyone makes when they conflate two things as the same - programming and business logic/knowledge (and I'd also throw in complex systems knowledge there too).

Often, understanding the code or modifying it is the easy part! I'm sure a decent amount of people on this website could master COBOL sufficiently to go through these systems to make changes to the code.

However, if I understand from my own career enough, knowing why those things are there, how it all fits together in the much broader (and vast) system, and the historical context behind all of that, is what knowledge is being lost, not the ability to literally write or understand COBOL.

Code is less and less the scares resource.... Good documentation is.
This makes no sense. If IBM supposedly gets a significant amount of revenue from COBOL (a dubious proposition) then wouldn't this actually help them as COBOL programmers are getting rarer and rarer?
Indian service companies can train some of their intake with COBOL, some obscure printer programming language, Clojure etc and give them anxiety about getting into a career dead end.
I have a close relative at one of the biggest COBOL shops in the US, and something tells me we're about to find out how we take the stability of our payments infrastructure for granted.

Their company no problem grinding older developers into retirement for the sake of padding their quarterly numbers, work-life balance is hell there. They refuse to try to compete with the modern developer market, senior level pay tops out around $125k. Despite what you may have read about experienced COBOL developer pay, know that is not the average experience. The talent pool was not replenished because they did not want to pay, overseas contracting firms also stopped training COBOL developers because their contractors could earn more building modern infra on AWS, so now they're between a rock and a hard place.

I have little doubt that we are going to see a massive payments infra failure as a result of this. Not because the AI is inherently bad, but because the promises of the tech combined with terrible management practices will create the perfect conditions for a catastrophe.

What's so difficult about COBOL anyway? How come in these conversations there's never any examples?
Nothing is so difficult about COBOL. It’s just old-fashioned and everything surrounding it is legacy. Most people seem to think it has a negative value when placed on a resume. Maybe that’s true, I don’t know.

I have trouble getting people to even look at C code these days. I don’t understand why devs are so afraid of old things.

I never in a million years thought I would see zero hedge on the front page of HN.

Times are a changin'

Luckily zerohedge links are automatically flagged, so you're not going to see it often.
I’ve always thought the whole point of staying in COBOL is not to make unnecessary changes, and that many required changes are critical and need experts who know how to handle them exactly.

There’s hardly any room remained for LLM.

Which language would you convert the COBOL to that has a compiler that compiles to the Z-series' fixed and floating decimal type machine instructions for financial calculations?