Failx

joined 1 year ago
[–] Failx@sh.itjust.works 15 points 1 month ago (1 children)

I applaud your ethics. But you don't know how close you are to falling from grace.

Just yesterday I had to remove perfectly tested, sensible and non-ai code from our production system, not because that it did not do what the author intended, but because what the author intended was flawed. And this is exactly what ai also cannot teach you right now: Taking a step back to realize that your code might be right, but your intentions are not.

Definetly keep at it. But be aware you will do the wrong things even with perfectly working code.

[–] Failx@sh.itjust.works 3 points 2 months ago* (last edited 2 months ago) (1 children)

In theory valid point and I approve but I'd disagree the sustainability factor. Even with renewables we won't have infinite energy, so we still need to think what we need to use the availabile energy for. The taining ist already ridiculously expensive and should the model become extremely popular it might still pull a lot of energy.

Needs to be evaluated on a per model usefulness / Energy Consumption basis.

And I doubt most current models would score very highly on that. (but please tell me those who do)

[–] Failx@sh.itjust.works 15 points 4 months ago (2 children)

Did I hear a "Rock and Stone"?

[–] Failx@sh.itjust.works 20 points 4 months ago

Most likely old folk(tm) I used Arch during the migration from init to systemd (2009-2013). Oh boy did things like to go boom at unforeseen moments. Like random segfaults after package updates, disappearing as suddenly as they started.

But admittedly after ~2013 Arch stabilised extremely compared to before. I remember having discussions around 2014 with people surprised that they didn't have update-introduced issues with their Arch install for 2 years at that time. Most of them never again until today.

Also: new users After aforementioned stabilisation period I got to know recently started Linux users who just did wired shit. Like accidentally deleting all kernel images on a Luks encrypted system or using unusual hardware which by chance Ubuntu or Fedora supported out of the box but would require kernel patches for Arch. They wanted to learn and they did learn but always perceived Arch as "more complicated" than the alternatives. But most of that was imo not the fault of Arch.

[–] Failx@sh.itjust.works 6 points 6 months ago (5 children)

Power BI is a different kind of beast though. Soon only available as part of Microsoft Fabric, a SaaS analysis platform. Sure, the technically inclined can use Python/R/Julia with MongoDB, a set of SQL DBs, some CI and Plotly/Dash, but that effectively requires to have some Software- and Data Engineers on staff and some dedicated machines/VMs. Power BI / Fabric is much cheaper for small to medium sized companies outside of IT.

[–] Failx@sh.itjust.works 3 points 1 year ago

The problem does not seem to be you logging into your Microsoft mail account, but logging into other mail servers:

"non-Microsoft accounts are synchronized with the Microsoft cloud"

i.e. by using outlook, you give your login data for your work/school/private mail accounts to a third party (Microsoft) which then pulls all your mail and calendars as a copy to their systems

Or: Your mail client is now a cloud app and most likely will at some point come with either a subscription or you handing over all your emails to Microsoft for advertising purposes