wizardbeard

joined 2 years ago
[–] wizardbeard@lemmy.dbzer0.com 1 points 5 days ago (1 children)

Sidestepped that completely. Got 64GB when I built my desktop because RAM prices were low.

20GB used up by whatever other shit I have open? No problem, still enough left for whatever I'm actively working on.

Suffice to say this has not actually helped with the issue.

Nostalgia is a hard drug

There's a lot more to this article than the summary blurb would indicate.

It's mainly talking about how regardless of actual quality of output, market forces around AI are now allowing manager types to require more output from "mid-upper" class workers, and it's all shifting those positions downward to being treated more like assembly line jobs than they have been for decades.

Concerning trends, driven largely by market forces instead of any true quality or capability of AI.

[–] wizardbeard@lemmy.dbzer0.com 16 points 6 days ago (2 children)

I've been through the hellscape where managers used missed metrics as evidence for why we didn't need increased headcount on an internal IT helpdesk.

That sort of fuckery is common when management gets the idea in their head that they can save money on people somehow without sacrificing output/quality.

I'm pretty certain they were trying to find an excuse to outsource us, as this was long before the LLM bubble we're in now.

[–] wizardbeard@lemmy.dbzer0.com 19 points 6 days ago

I kind of love that Kanye had that breakdown where he released the Heil Hitler track and went to twitter posting all caps shit saying he was a nazi, and the internet's collective reaction seems to have just kind of rolled their eyes and move on. Just pure apathy. It's not shocking or contreversial, more "Yeah, we all figured that out ages ago. Fuck off."

[–] wizardbeard@lemmy.dbzer0.com 34 points 1 week ago (1 children)

If it helps it make any more sense, from other posts of this comic on lemmy, the original posts seem to be on pixiv, which is on paper a site for hosting art, but in reality seems to exist primarily to host a lot of hentai. Mostly hentai.

I've also noticed a small trend lately where adult artists fish for comissions by making strangely horny webcomics.

[–] wizardbeard@lemmy.dbzer0.com 9 points 1 week ago (1 children)

It's hidden in the middle of the description, but this is DoomRL after being told they had to stop using the Doom IP as a fangame. Same person behind it, and my understanding is that it's even built off the same codebase (just updated a hell of a lot over the years). So less a "spiritual successor" and more a rebranding.

Which is awesome if you like DoomRL.

[–] wizardbeard@lemmy.dbzer0.com 32 points 1 week ago (1 children)

This is shockingly "public" considering there's legal proceedings now.

In another article or trailer it was explicitly stated that they worked with the community mod makers to package in all the "must have" texture and model mods. So it should look like what you previously could mod it into, lol.

I hope it does better than the last try.

In most public places in the US it is technically illegal to be publicly intoxicated. It's incredibly rare that anyone would do anything about it unless you're being a public nuisance or something though.

[–] wizardbeard@lemmy.dbzer0.com 14 points 1 week ago* (last edited 1 week ago)

Bunch of disparate thoughts about this. Apologies.


You can't just purity test an entire profession. Or at least it isn't reasonable to do so.

You have to dig, but there are artists and authors out there who have positive opinions about AI. I'm sure the folks more active on !techtakes@awful.systems could get a list together of their favorite sneer targets.


All the anti-ai tools are coming from programmers. anubis, iocaine, nepenthes, nightshade, and more.


I'd suggest that it's worth remembering that a large portion of any online discourse is going to be made up of teenagers, college students, and people new to whatever it is they're discussing. But almost all of them will talk with an implied authority on whatever subject it is anyway. Many subs on reddit have done demorgraphics polls (or polls that included a question about age) and the results makes the sillier places like relationship advice make a hell of a lot more sense.

I have almost a decade of work experience in IT, more if you count some intro to programming teaching gigs I did. AI coding assistance tools are positioned perfectly to be like crack for early learners, but don't actually add too much value for more experienced folks (especially not anything can't be done faster with existing non-ai tools, like an IDE's code snippets features for repeated boilerplate code, which is the most frequent excuse I see online).

That said, I have literally never met any dev in real life that thinks favorably of AI coding assistance.

What I see more and more is corporate management types being afraid that all the hype for AI coming out of Gartner etc means that if they don't force their employees to use it they will get left behind as their competitors magically produce more somehow. At that point as an employee your choices are to work to meet whatever shit metric they've constructed or give up making a paycheck.

It's very easy to tell other people they need to be more principled when it's not your food/shelter/insurance/livelyhood on the line.

And it's also worth noting that what the current wave of popular AI (LLMs) does best is generate text. So of course you're going to see text posts online trying to shift the window. That's the company using their horrific tool as designed. There was evidence of AI bots hyping AI in the programming subs on reddit that got overshadowed by them shutting off third party API access.


AI is only as inevitable for as long as the megacorps can keep funneling money into the pit. The world leader in AI conpanies, OpenAI, is doing so great with money that they are trying to threaten their business partner Microsoft out of roughly $40B. This could all blow up fairly soon. I hope it does.


There absolutely are devs out there who are getting ground down, where there principles are becoming eroded over time.

But like with many things in life, the rabble at the bottom aren't the ones effecting real long term change, media coverage, corporate adoption, etc. Don't turn this into some crab pot situation.

Go after the tech reporters continually giving AI the benefit of the doubt. Go after industry steering publications like Gartner. Go after politicians giving AI projects sweetheart deals that allow them to coast by without having to compete fairly based off their actual costs (OpenAI loses money on every request it serves). That allow AI datacenters to continuously violate local laws. That allow the datacenters to pollute their local water table. Go after the management enacting requirements that their subordinates demonstrate how AI is improving their work efficiency on pain of firing. You get the picture.

And sure, call out individuals. But please don't label us all as a group on this.


EDIT: old man intensifies. And another thing! (to support my point that the corps need to be the target in these discussions):

Companies like Microsoft are resorting to forcing AI features into their products and defaulting those AI features to ON in order to get the user numbers they need to keep justifying the absurd expenditure on AI, and to be able to keep pushing the false narrative into the media of everyone using AI.

These effectively fabricated increases in user numbers from these features being default on means that the impact of individual programmers and devs deciding to use or not use AI are not going to move the needle significantly enough to make a difference.

I'm proud to say that currently my workplace has all this shit disabled on an enterprise level, and firewall blocks preventing using any of it. Unfortunately that will only last as long as it takes for the suits to come up with a sufficiently paranoid acceptable use policy. We're doing what we can to keep stoking the idea that putting any info into these things is effectively selling it to potential competitors, but that can only go so far against Gartner and the like whispering in the C-suite's ear.

Anyway. Everyone has a duty to say no to this shit, but all the companies are already using every dirty trick they can to boost the user numbers. If you're running around under the impression that the absurd user numbers represent true users, and that those users must be comprised entirely of tech workers because artists never would use AI, then you're looking at a false narrative being pushed by the people who stand to make the most money off this shit. The user numbers are not accurate to begin with.

 

Probably need this disclaimer before half the shit I say.

 

Going way back to late 2000s internet memes with this one. "Ceiling cat watches you masturbate"

 
 

NIST is a US government org that releases industry guidlines on best practices for cybersecurity.

I know that infosec and sysadmin work aren't the same, but in my experience it often falls to sysadmins and systems engineers to fill the gaps. Hope this is useful.

 

NIST is a US government org that produces industry guidlines on best practices for cybersecurity, and they've just released a massive update to their framework.

 

Soichi Terada is a House music artist who was popular in Japan in the 90s. Outside of Japan, he's mostly known for his soundtrack work on the PS1 game Ape Escape.

This is one of his covers/arrangements/remixes, where he plays around with elements of another song. Not quite sure what to classify it as, otherwise I'd label it in the title.

I find his music to have a pretty distinct style, and I like using it as background while I study, code, or do other work.

 

I'm looking for a free, reputable ad blocker on the Play Store. Something that does local host/filter list filtering using the VPN feature, like Blokada 4 or 5 (before they started cloud hosting the filtering features as a money/data grab).

Personally, I'm no stranger to F-Droid or Obtanium and even have dipped my toes into ADB.

I need this for family members when they start asking, so I can point them at something decent that won't try to fleece them and get on with my life unburdened by family tech support hell. Something they can install through the Play Store they already have and easily switch on and off if something they "need" isn't working.

So that eliminates just setting their DNS to an ad blocking one in their Wi-Fi settings. Wouldn't follow them off that specific connection, and wouldn't be an easy toggle if something broke.

 

Microsoft's documentation for revoking user access from Azure AD currently references cmdlets from the AzureAD PowerShell module, which will be deprecated on June 30th.

Microsoft reccomends using the MSGraph module or API as a replacement for the AzureAD module, but I'm having a hell of a time with it.

I'm trying to figure out how to use PoweShell to wipe corporate data off a user's BYODs, and I'm stuck trying to get a list of a user's BYODs through Graph. Ultimately this will be part of automation kicked off when a user leaves the company.

Queries for devices and managed devices for a given user seem to be missing devices that are shown through Azure Portal when looking at a user in Azure AD and then looking at their devices. The query for deleting data is also unclear in whether it wipes the whole device or just corporate data.

Does anyone have any resources or guidance on this? Most of what I'm finding is outdated or too vague for me to be comfortable utilizing it.

view more: ‹ prev next ›