AI May Already Be Shrinking Entry-Level Jobs In Tech, New Research Suggests
- Reference: 0177839521
- News link: https://slashdot.org/story/25/05/28/2239206/ai-may-already-be-shrinking-entry-level-jobs-in-tech-new-research-suggests
- Source link:
> Researchers at SignalFire, a data-driven VC firm that tracks job movements of over 600 million employees and 80 million companies on LinkedIn, believe they may be seeing [1]first signs of AI's impact on hiring . When analyzing hiring trends, SignalFire noticed that tech companies recruited fewer recent college graduates in 2024 than they did in 2023. Meanwhile, tech companies, especially the top 15 Big Tech businesses, ramped up their hiring of experienced professionals. Specifically, SignalFire [2]found that Big Tech companies reduced the hiring of new graduates by 25% in 2024 compared to 2023. Meanwhile, graduate recruitment at startups decreased by 11% compared to the prior year. Although SignalFire wouldn't reveal exactly how many fewer grads were hired according to their data, a spokesperson told us it was thousands.
>
> While adoption of new AI tools might not fully explain the dip in recent grad hiring, Asher Bantock, SignalFire's head of research, says there's "convincing evidence" that AI is a significant contributing factor. Entry-level jobs are susceptible to automation because they often involve routine, low-risk tasks that generative AI handles well. AI's new coding, debugging, financial research, and software installation abilities could mean companies need fewer people to do that type of work. AI's ability to handle certain entry-level tasks means some jobs for new graduates could soon be obsolete. [...]
>
> Although AI's threat to low-skilled jobs is real, tech companies' need for experienced professionals is still rising. According to SignalFire's report, Big Tech companies increased hiring by 27% for professionals with two to five years of experience, while startups hired 14% more individuals in that same seniority range. A frustrating paradox emerges for recent graduates: They can't get hired without experience, but they can't get experience without being hired. While this dilemma is not new, Heather Doshay, SignalFire's people and talent partner, says it is considerably exacerbated by AI. Doshay's advice to new grads: master AI tools. "AI won't take your job if you're the one who's best at using it," she said.
[1] https://techcrunch.com/2025/05/27/ai-may-already-be-shrinking-entry-level-jobs-in-tech-new-research-suggests/
[2] https://www.signalfire.com/blog/signalfire-state-of-talent-report-2025
"job movements of over 600 million employees" (Score:4, Funny)
SignalFire, a data-driven VC firm that tracks job movements of over 600 million employees, over 27 of whom are real people.
Re: "job movements of over 600 million employees" (Score:1)
Hmm, so 599 999 973 AIs are looking for jobs? So many solutions looking for problems.
students: heres what to do (Score:3, Interesting)
Contribute to OSS for 2 years while you are still in school. Put this experience on your resume in the employment section. This way, the AI that replaced HR will sort your application with the other people who say they have 2 years of experience.
Re: (Score:2)
Why should this only count for AI? Experience is experience. And HR-AI will use the same stupid rules like other HR people ... if you're lucky with a bit less personal bias, but possibly even that changes when AI now starts to interpret images, audio and video to judge if you're a good fit.
Re: (Score:1)
That's a great way to get replaced by a cheap Indian.
I can tell you right now nobody gives a rat's ass what you did in school.
I know this because I just put a kid through school and I can tell you right now that's the last 2 years of school are composed of on the job training that you pay for now. I remember seeing my kids workload and being pretty fucking pissed off that I was literally paying for them to be trained on the job.
And when they got out of college it still took them 5 years to get t
Is it really, or is it Trump (Score:5, Interesting)
> Specifically, SignalFire found that Big Tech companies reduced the hiring of new graduates by 25% in 2024 compared to 2023
Incidentally, the threat of a Trump recession jumped dramatically over that time period.
> Asher Bantock, SignalFire's head of research, says there's "convincing evidence" that AI is a significant contributing factor
Convincing evidence would be doing a survey and asking companies why they reduced hiring. What Asher has done here is form a hypothesis (an important step in the scientific method, don't denigrate it, but not convincing evidence).
Re: (Score:2)
Probably both. When ecconomic stability goes away, companies only hire when they have to. Obviously, the orange moron and his deranged helpers do not even understand something this exceptionally obvious.
So it's kind of a catch 22 (Score:3)
During the Trump recession (The Trumpcession!) companies are going to fire you and trying to replace you with an AI.
I can't emphasize this enough, it does not matter if it works or not.
If it works great for them they get to replace you with an AI.
If it doesn't work the survivors will have to do double shifts to make up for the AI that doesn't work.
The real problem here is workers have absolutely no bargaining power whatsoever..
One thing I can say for sure automation has devastated the blu
Re: (Score:2)
> If it doesn't work the survivors will have to do double shifts to make up for the AI that doesn't work.
They don't have to.
What "entry level" jobs? (Score:2, Funny)
Those haven't existed since the 90s. If you don't have 20 years of experience at 20 years old you don't get a job these days.
Is it AI or overhiring? (Score:5, Informative)
I think when you look at the graph of software engineer jobs it looks far more likely that we are in a correction phase for overhiring:
[1]https://blog.pragmaticengineer... [pragmaticengineer.com]
I'm sure at some point there will be some effect of AI on entry level jobs, but in the companies where I stroll around I think the main issue is mostly that companies want to reduce the amount of programmers they have and not so much that AI is reducing the workload. I think a lot of companies would say it's because of AI because that looks better than most other explanations.
[1] https://blog.pragmaticengineer.com/software-engineer-jobs-five-year-low/
Re: (Score:2)
It's probably more complicated. Some of column a some of column b. Also don't forget everything we keep calling AI can just mean regular run of the mill automation.
There are two really nasty trends that folks don't like the think about.
The first is automation is real and it does take your job. In the past a lot of managers didn't want to automate White collar work. I had a manager for years who would order my team to do things by hand. I ignored them and that automation so that we could do other st
AI won't take your job? (Score:2)
The problem is, by and large jobs aren't taken - they're given by business owners and their MBA-driven middle managers.
If you're a business owner and you have a task that can be done in two ways:
* Human + AI = more productive human, human level salary
* All AI = OK performance, little or no salary
You're going to choose the second one until it becomes clear that "OK performance" is actively alienating your customers.
I predict many spectacular AI failures while this lesson is learned and relearned.
Re: (Score:3)
> I predict many spectacular AI failures while this lesson is learned and relearned.
Same here. And it will probably take several decades to fix the damage done. Because the other thing that is going to happen is that people will move to different fields for their education. Having a generation missing from an education field takes about 50 years to get fixed.
Well, good for all that still go into these areas, because "AI" will not cut it. I recently read a study that found that when selling insirance (something a lot more accessible to AI than general STEM jobs), AI saves a whopping 2.8% of
Re: (Score:2)
That's what will happen if AI levels off at the current level or just slightly above it. But there really are very few signs that the curve has started to level off. If it only doubles in competence over the next 5 years and then stops that forecast is going to be way off, and the signs are that it's improving faster than that.
Re: (Score:2)
Think the issue is that "OK performance" is largely out of reach of 'All' AI for most of the jobs people would be thinking of.
But assuming you get mediocre results, well, that's generally good enough. Slap a tray of dishes through a dishwasher. Sometimes they come through with stuff still stuck to the plates while a human dishwasher wouldn't make that mistake. Did that kill dishwashing machines? Not at all, still massively cheaper and if you *really* care a human can audit the results. Go to a buffet and ch
Re: (Score:2)
> If you're a business owner and you have a task that can be done in two ways:
> * Human + AI = more productive human, human level salary
> * All AI = OK performance, little or no salary
> You're going to choose the second one until it becomes clear that "OK performance" is actively alienating your customers.
It's also going to be funny when the people actually making the LLMs that give "OK performance" move past the user acquisition phase and put on their SaaS boots and start trying to cash in on their investments. If their LLM is worth the equivalent of Western knowledge-labor they aren't going to give unlimited queries for peanuts at some point. The calculus will get trickier for the business owner when it's tens of thousands of dollars for the "All AI" option.
Good luck with that (Score:2, Insightful)
I mean, how stupid can you be? How do these people think they are getting more senior people later?
Re:Good luck with that (Score:4, Interesting)
Yep. I've seen this in Canada for my whole life... nothing to do with AI... Employers have bleated that there isn't enough "qualified" potential hires... but they wouldn't put a dime into training. Yet, they never saw how that works. I thought it was a specific failing of Canadian ... lets call them "business leaders"... :-)
Re: (Score:2, Insightful)
Indeed. Personally, I can not even imagine being this extremely stupid. You have to willfully ignore the absolutely obvious.
Re: (Score:2)
Often the PHB attitude is: training is something that other companies do so that I can then hire their good people.
American healthcare system hurts too (Score:2)
As an American I cost around 10 to 15,000 a year or more to provide me health insurance on top of the premiums I pay which basically are an added expense to the company because they know they have to pay me enough to afford the premiums and the co-pays.
I have absolutely lost jobs because of this. Just recently I saw some work go to the UK that I was gunning for because they just cost a lot less than I do and it's not lower overall wages. Dollar for dollar in pure payout they cost the same however I cann
Re: (Score:2)
The answer is that AI is less the reason at this point and more the cover story. These companies had no idea what to do with the manpower and just kept hiring, because their competitors were hiring and they'll be damned if they look like idiots while competition hires. Hiring was a sign you had ideas, ambitions, commitments to your business plans and comfort that you have the financials to sustain it.
Now AI provides a new "cool" answer, a "good" company must have significantly offloaded their coding grun
Whose surprised? (Score:2)
The reality is a lot of basic tasks can be automated, and maybe not through AI, but how often have you written a tool to aggregate data, or, generate outputs that previously a person had to manually? I automate a ton of work and tasks, through medium-sized tools I've written for my company that replace data / IT monkeys. I don't need someone to sit on Azure and manually pull costs into a spreadsheet, I've automated it. Likewise, I don't need someone to stand up commonly used and configured Azure services
Getting experience without being hired (Score:2)
From TFS:
> A frustrating paradox emerges for recent graduates: They can't get hired without experience, but they can't get experience without being hired.
To get experience as a software engineer, all you need is a laptop (not even a great one), an internet connection (not even a fast one), and time. Most development tools are free, a lot of tutorials and documentation are free, and so are many libraries and other assets.
Create a GitHub profile. Make a public hobby project that goes beyond a todo list or space
BLS (Score:2)
BLS has recorded a growth of >20% in reported software developer since 2021. So, I'm not surprised a bunch can't find jobs. It doesn't really say anything about the industry, of course, just the pipeline of "candidates".
Re: Unsurprising (Score:3, Informative)
I'm kinda terrified to see AI generated assembly language.
Re: (Score:1)
Not worse than other code: Unreliable, impossible to maintain, insecure and generally crap. Oh, and as soon as we get liability for software (and we will), code like that will count as gross negligence and get your organization killed.
Re: (Score:3)
The tech giants would salivate for such a concept as it would effectively destroy open source software. Be careful about wishing for laws because they rarely have precisely and only the effect that the people enacting them desire. The big companies will be able to afford the legal teams and bribes to ultimately escape culpability whereas any smaller outfits won't.
Re: (Score:2)
> The tech giants would salivate for such a concept as it would effectively destroy open source software.
No, it would not. That is just a lie pushed but the COTS assholes to help them avoid liability. And you fell for the lie.
Re: (Score:2)
> I'm wondering whether we get to a point where compilers are obsolete, and AI just generates pre-compiled and optimized binary blobs.
I've seen this a few places and just don't see it happening. Obviously, there's the downside of just completely losing any hint of capability to audit/amend the intermediate coding language representation. So this has to be weighed against the upsides, of which no one seems to have any other than "it seems silly to have a machine generate output that is then transformed back to something for a machine". Just like the concept of a human hand-tuning assembly being able to outperform the best compilers is p