All skill degrade with disuse. For example, here in Canada we have observed a literacy and numeracy skills curve that peaks with post-secondary education and declines with retirement.[0]
That is one factor, but it’s not the whole thing. The other key element is “cognitive offloading” where your brain stops doing stuff when it thinks it is redundant.
This is similar to the photo-taking impairment effect where people will remember an event more poorly if they took photos at the event. Their brain basically subconsciously decides it doesn’t need to remember the event because the camera will remember the event instead.
If the tool is reliable, it's a win. Saved brain power doesn't disappear, it can be applied elsewhere.
If the tool is powerful enough to do a better job than our brains would, it's a big win. In fact, we built the entire technological civilization on one such fundamental tool: writing.
Or from another perspective: our brains excel at adapting to the environment we find ourselves in. The tools we build, the technology we create, are parts our environment.
This argument has held up in the past but there’s no certainty that during this current period where LLM’s are not perfect (and in many cases far from perfect) - they can ever become perfect that it’s fine for one’s existing human capital to depreciate.
I feel like it said a whole lot without giving me much to take action on. Like great, you summarized the current state of affairs but it doesn't make clear what I am to do about it.
> I feel like it said a whole lot without giving me much to take action on. Like great, you summarized the current state of affairs but it doesn't make clear what I am to do about it.
To be fair, not every article is a call to action. Sometimes they exist purely for informational purposes.
Are you making this claim specifically about this one particular post, or about everything on the blog, which dates back about a decade ?
Like "These Artemis 2 photos were generated by AI" is wrong but "The broadcast footage of the Apollo missions was generated by AI" is incredibly stupid and I want to understand if I'm about to engage with an incredibly stupid opinion.
There is a clear and sudden transition on this blog where prior to a certain date there are zero instances of the em-dash and then suddenly it appears like crazy. Like look at his archived posts from 2023, absolutely no em dashes... now look at every post from 2025 and almost every single one of them are literred with them.
The fact that the average person is seemingly incapable of detecting LLM text drives me insane. Every aspect of that article screams LLM. The tone, the punctuation, the sentence structure, the overall structure, it's so incredibly obvious. But the average person really is oblivious to it.
why?
Before comments about LLM I didn't notice this. After I compared pre-LLM posts and post-LLM and looks like AI was used to write/edit this article.
But.. why should I matter? Why my ignorance of this fact insane you?
It definitely feels like it was edited with AI, structure and tone seem like the usual AI "polish".
Lately I've been working on an embedded system with limited access to std lib/libc and this was very interesting but I was hoping to see more information, like how can I tell if something will be supported? what are alternatives? what does implementation specific mean?
I guess I am used to HN having more in-depth articles.
It's worth it. The increased participation and discussion have given a little momentum in usability, and AI on hand makes the learning curve very manageable. If you're already familiar with vmware, virtualization in general, it's a pretty easy transition.
I switched from VMWare to Proxmox a few years ago because Proxmox supported a wider range of network cards that were more common in the cheap desktop computers I use in my homelab, whilst VMWare almost required an Intel network card (which was usually fine for server hardware).
It was a surprisingly easy transition that I have not regretted one bit. I'm not sure whether there that was an actual migration path, without reinstalling servers from scratch. Homelab meant it didn't quite have the requirements of a production system...
I'm kinda hoping AI agents pass the threshold of being able to reliably do a complete production migration sometime this year. We've got a couple years left on the vmware contract, and it's just obscene what they've done, with the price hikes, degraded support, etc.
At this point they're more an enterprise scheme to maximize license profiteering for compatible software and OS "per core" licensing in conjunction with hardware platform providers. It'd be cool if the support were worth the price, but in most enterprise cases, you're going to save a lot of money if you pay enough full time staff to purchase, build, maintain, and operate a virtualization environment compared to what the enterprise platforms provide. In most cases, you'll save more than enough to keep a better specced system completely redundant with spare hardware on hand.
Honestly? VMs are a level of complexity I haven't felt a reason to fuss around with at home for at least the past five years. Just not interested.
I'm told that Kubevirt with Kubernetes has also been a winner among customers post Broadcom acquisition who were really reluctant to go beyond VMware previously.
Proxmox can do containers too and has other benefits like really good ZFS support. I only have a couple of VMs and everything else in containers on my little Proxmox server.
Proxmox can do LXC and has some experimental support for converting Docker based images... that said, it's not the same as Docker/Podman support, which are more feature rich.
I would suggest at least a minimal Linux Server VM if you're running containers, underneath ProxMox or on a bare metal install if you don't need other virtualization on said server.
There's still a great deal of Windows usage, but hopefully that will phase out with the passage of time. Canada's bureaucracy moves slowly, at the pace of generational attrition. It won't be until the last GenX retires that they could even meaningfully begin transitioning the average office worker away from Windows.
I work in government. Link 1 (2018) is essentially a dream. All of government got forced to use MS Dynamics CRM. Basically, anybody with a software requirement for case management, had to use MS Dynamics. I recommended we use Drupal in 2011. That was killed because everything had to be MS. I'm kind of surprised that it is in there given that nobody was allowed to use.
Link 0 and 2 are essentially from TBS and CDS. They coexist together. They are essentially working at the very top as entities that gather information from other departments. They can do whatever they want because they help write the rules.
I'm not trying to discredit your post, just saying that as someone who has brought OSS tools to development at the government and tried to use OSS tools for client (I failed at that), it is nearly impossible at the moment. We are married to Microsoft and its cloud.
I do agree, that it may take an entire generation because right now, 190+ departments are not exactly jumping to FOSS, and in many situations, they are down right told you are not allowed.
In addition, the current de facto document management system is from OpenText. Although many just use Sharepoint Online.
Ironically, as everything moves to the cloud, it would be easier to move to a solution that is FOSS based, and based in the cloud. Technology has matured enough that you don't need executables on a desktop, you just need a browser pointing to a website.
We use Microsoft Dynamics 365 (model-driven app) at work, it's rarely mentioned on HN and people don't know how insanely bad this P.O.S. software is.
From the botched implementations of AG Grid to their crippled version of CKEditor (with Copilot forced in of course), the daily bugs are an absolute nightmare.
And then most support tickets (if you can even open one after a forced chat session with Copilot), get handled by a third-party, most likely in India with different timezones than you and the support calls are a crapshoot.
The Phoenix contract predates the more recent efforts to switch to FOSS.
But also, Canada loves to burn money on American suppliers. It's probably why the recent interest in _Buy Canadian_ has the American administration annoyed.
Phoenix was a literal trap laid by the Conservative government just before leaving knowing it would be a shit show for the Liberals in the coming years.
I was part of a SaaS company of diehard GenX Windows fans.
Decades of abuse by Microsoft has definitely hurt them: they have lost hope and are cynical about the future of Windows. I reckon they would switch away if they could afford to.
Every year Microsoft does something to make you feel like you're being screwed over.
We only just missed taking a silverlight bullet. Windows phone wasted over a year of development. Internet Explorer doubled development costs. The OS version churn is expensive. However SQL server has been a good foundation.
Microsoft used to love developers. They just abuse them now. Even Apple is nicer to developers!
I find it's often faster for me to finish the final 20% myself than to talk the agent into doing it for me; because too often the agent will start to eat its own tail, and spend far too long completing something that I find obvious.
Yes, and English/natural language is not necessarily more concise than programming languages, if you need to describe something precisely.
For example, I was recently trying to get an agent to debug something which was difficult to debug because it ran in an exotic context, where debuggers and logging and printf couldn't easily reach. The agent kept coming up with more and more elaborate and smart-sounding theories and debugging strategies, but nothing worked. I stupidly kept going with this for like 20 minutes, until finally I just went into an IDE, did a simple "comment bisection" where I commented stuff out until I found the line that was breaking, and found and fixed the problem in five minutes. So I solved it by typing code. The code I typed: "//" (in about six places). I could probably have gotten the agent to do the same thing but would have actually literally had to type more to explain to the agent what I wanted. In fact it took me longer to write this comment describing what I did here than it did to just do it.
> You're spending more time orchestrating than creating.
Orchestration is a form of creating. I've lead teams of programmers; while it is different than orchestrating AI, programmers typically require less hand-holding, it is not so different in how it is a form of delegating effort to achieve your creative goals.
> The agents aren't the problem. Your brain is.
If anything, my worry is that relying too heavily on agents will cause my knowledge to be forgotten and my skills to atrophy. I don't particularly want to stop programming so much as it is that I want to develop software as part of a team. That team now includes some AI agents, as well as humans.
The need to write code isn't going anywhere. I expect that in the very long term it will retain value, as developing expert level programming ability will be a difficult challenge when so much can be accomplished with little to no such ability.
I'm not looking for instructions about how to get Barbie Riding Club to work. Our conversation led to a vague memory of the blog post/release notes. I'm looking for that blog post, the list of titles, and the short subthread about it that I mentioned not being able to find.
It's apparently explicitly supported by dxwnd, but I failed to find anything in release notes or anything like a post you mention. Binary download however contains an 'exports' directory which is basically a list of titles it supports.
Great reading in itself.
Use it or lose it, as it were.
0: https://www150.statcan.gc.ca/n1/daily-quotidien/241210/dq241...
reply