@Yet another Josh :donor:
#Sysad. #SigInt. #Hacker. #Pagan. #musician. #renaissance.
In the #SCA I'm Vincent O'Cosnachain of the Midrealm
Relays (1)
- wss://relay.ditto.pub – read & write
Recent Notes
Which Linux distros have officially taken a stand against AI in their OS?
#Linux #NoAI
@nprofile1q... so, a "no tools" policy?
LLMs are a tool. And if people are NOT being responsible for their tools, then its a person problem.
Blaming a large vector array is pretty dumb, tbh.
LLMs are a tool. And if people are NOT being responsible for their tools, then its a person problem.
Blaming a large vector array is pretty dumb, tbh.
the infosec people at my work are rioting because the Distant Corporate Overlord sent an email that scores 10/10 on the phishing scale (“We want to give you a present to thank you for all your hard ...
@nprofile1q...
The real scary email isnt some dumb phishing. The scary is straight forward.
"We are a #ransomware operator. We would like for you to run this script on your work machine. If you do, we'll pay you $1000 in your choice of crypto. If they pay the ransom, we pay you 10%."
That weaponizes ransomware so that everybody is a potential #insiderthreat. And given these days with so much job abuse due to terrible conditions, sending a few of these emails are sure to hit someone disgruntled enough to say fuckit.
The real scary email isnt some dumb phishing. The scary is straight forward.
"We are a #ransomware operator. We would like for you to run this script on your work machine. If you do, we'll pay you $1000 in your choice of crypto. If they pay the ransom, we pay you 10%."
That weaponizes ransomware so that everybody is a potential #insiderthreat. And given these days with so much job abuse due to terrible conditions, sending a few of these emails are sure to hit someone disgruntled enough to say fuckit.
"Please set a new password for your account."
okay
"Your password does not meet your organization policy."
okay, you could have hinted that in the interface before I hit enter, but what's the polic...
@nprofile1q...
Password(month)(year)!
PasswordMay2026!
Yeahhhh, that works mostly everywhere, easily updates, and is a totes terrible password.
But most orgs cant even be arsed to follow NIST's 2017 password directive, let alone 2025's.
Password(month)(year)!
PasswordMay2026!
Yeahhhh, that works mostly everywhere, easily updates, and is a totes terrible password.
But most orgs cant even be arsed to follow NIST's 2017 password directive, let alone 2025's.
@nprofile1q... @nprofile1q...
Your "GOTCHA" style prose is your downfall, especially in reaching the normies and people dabbling with LLMs.
Of course I know not to trust it. It isn't about encouraging smart people from not using LLMs. Its about the 100 IQ and lower from not using them.
The "joke" about 90% correct is that they are wrong 10% of the time. Good fucking luck figuring WHAT 10%.. Unless you're an expert in that area, it'll hallucinate and lie and you wont have a bloody clue.
And yea, they do seem to know everything. And they dont.
The ONLY answer isnt "make fun of everyone who uses an LLM". The answer is to teach what these tools are doing, and why, and where they dont work. Its my belief that when people start studying the patterns, that they too will see the deficiencies. Doubly so if you teach someone to probe an LLM over a topic they know well. It'll g quickly from 'citing general facts' to 'making shit up authoritatively'.
Through education, people will realize what these things are, and how they fail.
Shaming/insulting them for using LLMs will only make them quit talking to you about it.
https://www.youtube.com/watch?v=y85nqc2zm7M talk exactly about how the whole abstinence recommendation never ever fucking works. Nor does shaming. Knowledge and education does work, every time.
Your "GOTCHA" style prose is your downfall, especially in reaching the normies and people dabbling with LLMs.
Of course I know not to trust it. It isn't about encouraging smart people from not using LLMs. Its about the 100 IQ and lower from not using them.
The "joke" about 90% correct is that they are wrong 10% of the time. Good fucking luck figuring WHAT 10%.. Unless you're an expert in that area, it'll hallucinate and lie and you wont have a bloody clue.
And yea, they do seem to know everything. And they dont.
The ONLY answer isnt "make fun of everyone who uses an LLM". The answer is to teach what these tools are doing, and why, and where they dont work. Its my belief that when people start studying the patterns, that they too will see the deficiencies. Doubly so if you teach someone to probe an LLM over a topic they know well. It'll g quickly from 'citing general facts' to 'making shit up authoritatively'.
Through education, people will realize what these things are, and how they fail.
Shaming/insulting them for using LLMs will only make them quit talking to you about it.
https://www.youtube.com/watch?v=y85nqc2zm7M talk exactly about how the whole abstinence recommendation never ever fucking works. Nor does shaming. Knowledge and education does work, every time.
The constant mental vigilance in a generative world is exhausting.
"I asked Claude to do $thing and it did this!"
No it didn't. No you didn't. Probably none of that happened.
And somehow, being unw...
@nprofile1q...
Thats what you and most AI haters miss.
LLMs are right like 90% or so at a time.
Even 80B models are right a big majority of the time. This is why normies and managerial staff are like "this is amazing". They dont know what they dont know.
Its that last 10% that breaks all sorts of stuff and people. And you have to be a specialist in that 10% to know when its bullshitting/hallucinating/lying to you.
Thats what you and most AI haters miss.
LLMs are right like 90% or so at a time.
Even 80B models are right a big majority of the time. This is why normies and managerial staff are like "this is amazing". They dont know what they dont know.
Its that last 10% that breaks all sorts of stuff and people. And you have to be a specialist in that 10% to know when its bullshitting/hallucinating/lying to you.
@nprofile1q...
"The one weird thing about cats is that 'murder mode' isnt the first photo, but tge second one:"
"The one weird thing about cats is that 'murder mode' isnt the first photo, but tge second one:"
@nprofile1q...
I just had this sent over youtube shorts.
My guess is some LLM spammer got it. But wanted to ping you
https://www.youtube.com/shorts/C20z7EMMouY
I just had this sent over youtube shorts.
My guess is some LLM spammer got it. But wanted to ping you
https://www.youtube.com/shorts/C20z7EMMouY
I've been building out my repetoir of irish music so I can play in local sessions. I'm working on the common ones first. So far I've memorized the "Swallowtail Jig" and I'm currently working on memori...
@nprofile1q... theres also songs that use the other spelling "jegg" or "jegge".
1651 Playford manuscript has quite a few, like Lord Caranavan's Jegg, Kemp's Jegg, Millison's Jegge.
1651 Playford manuscript has quite a few, like Lord Caranavan's Jegg, Kemp's Jegg, Millison's Jegge.
1
So Rachel Reeves has confirmed the austerity mindset continues in Govt., by announcing that the increase of defence budgets warranted (to be fair) by the current situation in the Middle East & the con...
@nprofile1q...
@nprofile1q... has a great article he published 3 days ago. Austerity politics creates fascism.
https://doctorow.medium.com/https-pluralistic-net-2026-04-13-always-great-our-nhs-f612bbaa0d4a
@nprofile1q... has a great article he published 3 days ago. Austerity politics creates fascism.
https://doctorow.medium.com/https-pluralistic-net-2026-04-13-always-great-our-nhs-f612bbaa0d4a
"On the acceptance of GenAI"
https://smallsheds.garden/blog/2026/on-the-acceptance-of-genai/
https://tldr.nettime.org/system_nettime/media_attachments/files/116/369/383/660/367/551/original/0c4cfd13c...
@nprofile1q...
None of these are true if you run your own LLMs on your own hardware, using FLOSS models.
But the #MastodonHOA has deemed all AI to be abhorrent as a blanket decision.
None of these are true if you run your own LLMs on your own hardware, using FLOSS models.
But the #MastodonHOA has deemed all AI to be abhorrent as a blanket decision.
1
nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqczqe8hpgugz54l4mqeqnsqld7rzzcv2hl4j8kx8lzu6c7uhtj6mq99stq7
I'm really glad my mom isn't on the fedi... I could still get in a lot of trouble fo...
@nprofile1q... @nprofile1q...
Honestly, your mom sounds like a hyper-religious piece of crap.
I knew only a scant few who had to hide TVs. They were the hyper-christian-fu damentalists where everything was the devil.
Fuck all of those types.
Honestly, your mom sounds like a hyper-religious piece of crap.
I knew only a scant few who had to hide TVs. They were the hyper-christian-fu damentalists where everything was the devil.
Fuck all of those types.
1