The title is a bit misleading. Reading the article, the argument seems to be that entry-level applicants (are expected to) have the highest AI literacy, so they want them to drive AI adoption.
They hire juniors, give them Claude Code and some specs and save a mid/senior devs salary. I believe coding is over for SWE's by end of 2027, but will take time to diffuse though the economy hence still need some cheap labour for a few years, given the H1-B ban this is one way without offshoring.
Customer interaction has imo always been one of the most important parts in good engineering organizations. Delegating that to Product Managers adds unnecessary friction.
Why is that bad? You write better code when you actually understand the business domain and the requirement. It's much easier to understand it when you get it direct from the source than filtered down through dozens of product managers and JIRA tickets.
Not sure why this is being downvoted. It’s spot on imo. Engineers who don’t want to understand the domain and the customers won’t be as effective in an engineering organization as those who do.
It always baffles me when someone wants to only think about the code as if it exists in a vacuum. (Although for junior engineers it’s a bit more acceptable than for senior engineers).
Another one? What is it with IBM, they must really save lots of money in a way no one else has figured out by firing people at 50yo. This is like the 3rd or 4th one i've heard from them.
The title could be dead wrong; the tripling of junior jobs might not be due to the limits of AI, but because of AI increasing the productivity of juniors to that of a mid or senior (or at least 2-3x-ing the output of juniors), thus making hiring juniors an appealing prospect to increase the company's output relative to competitors who aren't hiring in response to AI tech improvements. Hope this is the case and hope it happens across broadly across the economy. While the gutter press fear mongers of job losses, if AI makes the average employee much more useful (even if its via newly created roles), it's conceivable there's a jobs/salaries boom, including among those who 'lose their job' and move into a new one!
Interesting signal from IBM. The "AI will replace all junior devs" narrative never accounted for the fact that you still need humans who understand the business domain, can ask the right questions, and can catch when the AI is confidently wrong. Turns out institutional knowledge doesn't just materialize from a model — you need people learning on the job to build it.
Perhaps I'm being cynical, but could they be leaving out some detail? Perhaps they're replacing even more older workers with entry level workers than before? Maybe the AI makes the entry level workers just as good-- and much cheaper.
Not because it's wrong, but because it risks initiating the collapse of the AI bubble and the whole "AI is gonna replace all skilled work, any day now, just give us another billion".
To a non-technical individual IBM is still seen as a reputable brand (their consulting business would've been bankrupt long ago otherwise) and they will absolutely pay attention.
The title is a bit misleading. Reading the article, the argument seems to be that entry-level applicants (are expected to) have the highest AI literacy, so they want them to drive AI adoption.
They hire juniors, give them Claude Code and some specs and save a mid/senior devs salary. I believe coding is over for SWE's by end of 2027, but will take time to diffuse though the economy hence still need some cheap labour for a few years, given the H1-B ban this is one way without offshoring.
IBM has cut ~8,000 jobs in the past year or so.
Sounds like business as usual to me, with a little sensationalization.
"software engineers will spend less time on routine coding—and more on interacting with customers"
Ahh, what could possibly go wrong!
Customer interaction has imo always been one of the most important parts in good engineering organizations. Delegating that to Product Managers adds unnecessary friction.
Why is that bad? You write better code when you actually understand the business domain and the requirement. It's much easier to understand it when you get it direct from the source than filtered down through dozens of product managers and JIRA tickets.
Programmers have an unfortunate tendancy to be too honest!
Not sure why this is being downvoted. It’s spot on imo. Engineers who don’t want to understand the domain and the customers won’t be as effective in an engineering organization as those who do.
It always baffles me when someone wants to only think about the code as if it exists in a vacuum. (Although for junior engineers it’s a bit more acceptable than for senior engineers).
Sounds like we're finally doing agile.
I’m a people person.
https://www.youtube.com/watch?v=hNuu9CpdjIo
Interesting given the current age discrimination lawsuit:
https://www.cohenmilstein.com/case-study/ibm-age-discriminat...
Another one? What is it with IBM, they must really save lots of money in a way no one else has figured out by firing people at 50yo. This is like the 3rd or 4th one i've heard from them.
The title could be dead wrong; the tripling of junior jobs might not be due to the limits of AI, but because of AI increasing the productivity of juniors to that of a mid or senior (or at least 2-3x-ing the output of juniors), thus making hiring juniors an appealing prospect to increase the company's output relative to competitors who aren't hiring in response to AI tech improvements. Hope this is the case and hope it happens across broadly across the economy. While the gutter press fear mongers of job losses, if AI makes the average employee much more useful (even if its via newly created roles), it's conceivable there's a jobs/salaries boom, including among those who 'lose their job' and move into a new one!
Interesting signal from IBM. The "AI will replace all junior devs" narrative never accounted for the fact that you still need humans who understand the business domain, can ask the right questions, and can catch when the AI is confidently wrong. Turns out institutional knowledge doesn't just materialize from a model — you need people learning on the job to build it.
Probably not on the IBM jobs site yet, where the number of entry level jobs is low compared to the size of the company (~250k):
https://www.ibm.com/careers/search?field_keyword_18[0]=Entry...
Total: 240
United States: 25
India: 29
Canada: 15
Aren't those general jobs opening. Like junior swe only needs a single generic posting for all positions
Perhaps I'm being cynical, but could they be leaving out some detail? Perhaps they're replacing even more older workers with entry level workers than before? Maybe the AI makes the entry level workers just as good-- and much cheaper.
Bold move.
Not because it's wrong, but because it risks initiating the collapse of the AI bubble and the whole "AI is gonna replace all skilled work, any day now, just give us another billion".
Seems like IBM can no longer wait for that day.
Is IBM invested big in LLMs? I don't get the impression they have much to lose there.
Their CEO already said what he's thinking about all the spending [0].
[0]: https://news.ycombinator.com/item?id=46124324
Good. Nobody needs to rip that bandaid off. Might as well be IBM.
I mean it’s IBM. On average, 70% of their decisions are bad ones. Not sure I’d pay a single bit of attention to what they do.
To a non-technical individual IBM is still seen as a reputable brand (their consulting business would've been bankrupt long ago otherwise) and they will absolutely pay attention.
Yeah, they are only 114 years old. How they can have the knowledge to stay afloat in trying times like this?
[dupe] Earlier: https://news.ycombinator.com/item?id=46995146
Thanks - we-ve merged that thread hither.