←back to thread

The great displacement is already well underway?

(shawnfromportland.substack.com)
511 points JSLegendDev | 1 comments | | HN request time: 0.208s | source
Show context
JohnMakin ◴[] No.43976144[source]
I’m not trying to be unsympathetic in this comment so please do not read it that way, and I’m aware having spent most of my career in cloud infrastructure that I am usually in high demand regardless of market forces - but this just does not make sense to me. If I ever got to the point where i was even in high dozens of applications without any hits, I’d take a serious look at my approach. Trying the same thing hundreds of times without any movement feels insane to me. I believe accounts like this, because why make it up? as other commenters have noted there may be other factors at play.

I just wholly disagree with the conclusion that this is a common situation brought by AI. AI coding simply isnt there to start replacing people with 20 years of experience unless your experience is obsolete or irrelevant in today’s market.

I’m about 10 years into my career and I constantly have to learn new technology to stay relevant. I’d be really curious what this person has spent the majority of their career working on, because something tells me it’d provide insight to whatever is going on here.

again not trying to be dismissive, but even with my fairly unimpressive resume I can get at least 1st round calls fairly easily, and my colleagues that write actual software all report similar. companies definitely are being more picky, but if your issue is that you’re not even being contacted, I’d seriously question your approach. They kind of get at the problem a little by stating they “wont use a ton of AI buzzwords.” Like, ok? But you can also be smart about knowing how these screeners work and play the game a little. Or you can do doordash. personally I’d prefer the former to the latter.

Also find it odd that 20 years of experience hasnt led to a bunch of connections that would assist in a job search - my meager network has been where I’ve found most of my work so far.

replies(20): >>43976259 #>>43976303 #>>43976353 #>>43976380 #>>43976637 #>>43976752 #>>43977119 #>>43977214 #>>43977247 #>>43977438 #>>43977464 #>>43977541 #>>43977653 #>>43977728 #>>43977806 #>>43977828 #>>43977966 #>>43978694 #>>43979741 #>>43981729 #
bradgessler ◴[] No.43977214[source]
It feels like we're in a phase where hiring is slow for a lot of reasons:

1. Lot's of great talent on the market. It's a great time to be owning a company right now in terms of hiring.

2. The reality and perception of AI making it possible to do "more with less". I can imagine conversations playing out today, "we need to hire more developers" with the rebuttal, "ok, what about AI? Let's see how far it will go without hiring more people"

3. Even without AI, software teams can do more with less because there's simply much better tooling and less investment is required to get software off the ground.

4. Interest rates and money is simply more expensive than it was 3-5 years ago, so projects need to show greater return for less money.

It does feel like the reality and perception of AI hasn't converged yet. There's a general sense of optimism that AI will solve a lot of huge problems, but we don't really know until it plays out. If you believe history rhymes, humans will figure out what AI does well and doesn't do so well. Once that's worked out, the gap between perception and reality will close and labor markets will tighten up around the new norm.

replies(4): >>43977575 #>>43978377 #>>43978894 #>>43987085 #
1. araes ◴[] No.43978377[source]
> It does feel like the reality and perception of AI hasn't converged yet

I learned a word cruising Reddit the other day that summarizes that issue quite well - "liminal". At the time, it was in the context of malls, and the collapse of American storefront consumerism, yet the issues are similar:

  "relating to the transitional stage of a process", or "the quality of ambiguity or disorientation that occurs in the middle stage"
> general sense ..., but we don't really know ... the gap between perception and reality will close and labor markets will tighten up

We're stuck in that in-between land where your 2) seems like it's often the response to most suggestions. We'll, we don't really want to take a risk ... cause tomorrow AI may make that choice irrelevant. We don't really want to invest ... cause tomorrow AI may make our investment worthless. We don't really want to hire more people ... cause tomorrow AI may do their jobs easily. And there's always that number 3) sensation somewhere "your team can do more, you're just not leveraging tools enough".