Ask HN: Many SWEngs out there are considering re-training
So much so that, a la industrial revolution, there will be lots of layoffs and SWEngs without a cause. This is obviously quite a pessimistic view of where the industry is going, and maybe I'm just looking for some reassurance or discussion - my question is at what point do a large amount of SWEngs have to find something else to do and re-train?
ai or not, i've always wanted to start my own business, and ai likely will make that easier (maybe)
Software contractors/freelancers are going to be in a difficult spot. Before it was their convenience that allowed for a higher hourly rate, but now it's just as convenient to throw your pre-existing team at the problem - as not having knowledge of a specific tech stack is no longer as prohibitive as it once was
You can run agents by yourself on your 9-5 and still make crucial decisions in software industry.
The other companies will stay where they are at right now.
Learning to manage code bots is way easier if you know code. Managing code bots when you don’t know code leads to how SWEs have been mismanaged in the past.
Not a junior either, but I'd rather do anything else than be expected to turn into a middle manager for AI agents. That's not why I got into computers.
I disagree. Throughout the history of computing the number of SWE's has gone up and they have become, overall, more capable [0]. That hasn't cause a need for fewer of us, in fact the demand remains overall strong.
I predict instead it will drive the need for more of us rather than less. And potentially even "better" (or at least healthier as I see it) jobs. What I mean by that is, instead of outsourcing (locally or abroad) companies will be able to bring someone in-house. In the past, dropping the load on 1-2 people would have been untenable, with LLMs/AI a number of things become plausible. And this scales to every level. The company that previous outsourced their development and could have hired 3-5 developers (but that wasn't enough) can now consider hiring them.
Yes, the larger corporations might decide they can get the same done with fewer employees, or they might decide to do more with the same.
I do think a number of SaaS tools might be in more trouble if companies decide to just build in-house alternatives that are custom for their exact needs. But perhaps SaaS will adapt to "source available" or some other kind of way to allow easier extension of their core offering so that companies just build a sidecar or thin layer over the core that covers their needs better than a "one size fits all" SaaS ever could.
Perhaps this is too rosey of a view of the future but it's one that I think is at least partly backed by the history of the industry.
[0] And at least what they're able to produce, even if you disagree about the quality.
Very good point, something that I've encountered at my company already.
On the receiving end, as in losing customers to this? Or is your company replacing tools they used to pay for?
I'm very interested in how other places are handling this.
Another interesting one is the hosting of the open source equivalents of SaaS products on on-prem hardware. Turns out that agents take very well to infrastructure-as-code setups, making managing deployments of software far less work than it used to be. So we can use X database (albeit an older and less feature-rich version) on our own infrastructure that we can easily scale/tailor to our needs without paying for the enterprise SaaS version
I hope that open source software and/or slightly modified open source software sees a huge surge because of this. A move to "open core" products (free or paid) that then companies argument/extend as they need for their specific use-case.
Think about people who left the industry in the dot-com era and how that probably seemed smart at the time
None of these ideas are as lucrative as writing SW. I'm not technically trained as a SWE though I do write it, I'm a hardware person.
One good engineer plus AI > one good product owner plus AI. The former can build, the latter can’t (unless we are talking about very simple toy products or prototypes)
Unless AI automates SWE completely. Which I don't think will happen. If AI is that powerful, it would be a completely different world.
Though I think there are some things that are true right now.
* The way SWEs work is changing a lot (though I think it's just a progress that always happened) and a lot of people might not like it
* Currently there is a lot of uncertainty and FOMO in a lot of companies, how to operate, what to prioritise. Hence it might be not a great time, sometimes there are decisions like layoffs, not hiring junior developers, etc. I think that stabilises with time.
This is stated main goal many are persueing at this point.