Paco Hope

Amateur professional #selfhost sysadmin. Professional amateur #cloud #security at #AWS. Also fond of #cats, #cigars, #whiskey and #pipes. I like board games and some video games. I am #covid cautious and I still #wearamask. Opinions are my own, but they can be yours too. 100% Organic:,No artificial colors, preservatives, or intelligence added.

Public Key

npub1wemf3zlwy2dra8tfeq0my8fl3hk3k35pr7vppsua2qj58yf8v0vq88epgj

NIP-05 Address

paco@infosec-exchange.mostr.pub

Profile Code

nprofile1qqs8va5c30hz9x37n45us8ajr5lcmmgmg6q3lxqscww4qf2rjynk8kqpz4mhxue69uhhyetvv9ujuerfw36x7tnsw43qz9thwden5te0wfjkccte9ekk7um5wgh8qatz02nm78

Publishing to

relay.ditto.pub

relay.mostr.pub

Author Public Key

npub1wemf3zlwy2dra8tfeq0my8fl3hk3k35pr7vppsua2qj58yf8v0vq88epgj

Show more details

Last Notes

2026-10-26 23:01:55 UTC

Here’s my edit of the Marisa Mayer quote. Bracketed words are mine.
“A lot of what excites me is taking something that used to [be meaningful because it would] take you hours or minutes, and now [stripping it of human connection and] getting it done in seconds,” Mayer told me at her company’s office in Palo Alto last month.
The thing is, all her examples aren’t things that are time consuming because they’re tedious and low-value. Her examples are time-consuming because they take thought and care. The 3 examples listed in the free chunk of the NYT article are:
writing 3 positive qualities about her own fucking daughter for a teacher
shopping for a pair of pants she took a photo of
choosing a gift for a picky spouse
I really have a hard time being sympathetic to someone who skips the hard work of 1 and 3 and reaches for AI.
You get good at what you practice. If you practice avoiding these things, you’ll get good at avoiding these things.
https://www.nytimes.com/2026/10/25/style/marissa-mayer-ai-assistants-women.html?unlocked_article_code=1.LlE.Pd6S.asvWuB4qCloE&smid=url-share

2026-10-26 00:33:17 UTC

I've offered to spend time reading stories to my friend's 10-year-old via audio/video conference. She tells me that her daughter would really be excited to do it from her own phone (i.e., not using mom's phone). That's fine, I said, but you'll need to install Signal. I don't use any of the other bullshit.
So now we will have a 10-year-old using Signal. That's how we do it. One person at a time.

2026-10-26 00:22:37 UTC

Hey #Linux folks using laptops. I have installed #debian on an Apple MacBook circa 2015. When this thing ran MacOS, I could close the lid with a full battery and it would stay suspended for days at a time. Now that I'm running Linux on it, it runs the battery down in hours (less than 24) when I suspend.
What's the trick? I'm using KDE Plasma, if that matters.
Please don't give any suggestions that:
Suggest a different Linux distro. I'm not interested.
Suggest not using KDE. That's not the question.
Recommend something you've never tried on Debian/Ubuntu or one of those variants.

2026-10-25 11:37:32 UTC

I wonder which took longer? Counting 28 teeth on a sprocket in a photo or using the LLM that makes us so efficient and more productive?
I know which approach burned the most non-renewable energy sources.
https://mastodon.nl/@jap/117495812146047462

2026-10-19 01:33:33 UTC

Paint me like one of your French Zombies.
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/464/996/010/367/558/original/6edb76704659befc.jpg

2026-10-17 12:50:59 UTC

Connecting dental floss to the internet was a mistake.
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/456/332/917/063/459/original/ae9736f83012b211.jpeg

2026-10-11 21:21:41 UTC

So in my experience getting #debian onto my Macbook 2015 ended up without any encryption on the local disk. Performance with LVM was abyssmal. Installing with ZFS on root left me with an unusable system. Something about the way debbootstrap installed during the instructions left me with a version of Linux that could not access the Apple USB devices. It would boot and the kernel would throw a bunch of USB device errors, and it wouldn't recogize the built-in keyboard, so I couldn't even login.
A standard Debian install works fine.
So #debian13 is running just fine. But this is a slow machine. I discovered a tool called eatmydata. It basically shims calls to fsync() system calls and does a no-op instead. (😱) What could possibly go wrong?
But running sudo eatmydata apt install blah blah blah runs way faster than the equivalent apt command.
It has breathed new life into this laptop.

2026-10-09 13:50:40 UTC

Long tech story with a happy ending and a funny punchline. I'll relegate some details to the replies.
I have an old (2015) MacBook. Something got corrupted. I can get to the drive in the recovery OS, but it won't boot. No problem. I was thinking of putting debian on it anyway. I figure I'll boot to recovery, get the data off it, reformat, Bob's your uncle.
...skipping how I wrecked the recovery OS and got it to where it wouldn't boot at all...
Target disk mode to the rescue! You can boot a mac of this vintage, press the letter T during boot, and it comes up like it's an external drive. But it's thunderbolt. And all modern macs are USB-C.
...$120 in adapters and cables later...
My 2021-vintage mac can see the 2015-vintage mac as an external drive! It prompts me for the FileVault password. It doesn't work. I KNOW my password. I've typed it daily for the entire 11 years I've used this laptop. Oh shit. The data is right there but it's encrypted.
My wife has an account on my laptop and I know her password. I try it. No dice. Oh double shit. Am I really not going to get my data? I'm suspecting corruption in the FileVault keys/data.
10 years ago, when my son was 7 I put an account on here for him so that he could play some game on it. The password, when he was that age, was simply spiderman.
I type spiderman and it decrypts! The drive mounts! fsck runs with no errors! I'm recovering data off the laptop! The progress bar on the rsync command could easily be showing my blood pressure going down.
#sysadmin #laptop #techsupport

2026-10-08 23:50:59 UTC

Seems to me that “backfiring squads” is what we have now. And they’re not very entertaining.
https://theonion.com/doj-to-bring-back-firing-squad-as-means-of-entertainment/

2026-10-07 02:51:54 UTC

My son game me the summary of The Hobbit as “Unexpected item in Baggins area” and now I can’t stop thinking about it.
It’s not his original thought. But it is hysterical.

2026-10-06 22:25:48 UTC

Yesterday I found $10 in a winter coat pocket. I’m going to multiply that by 365 and declare my closet’s annual run rate to be $3650. Anybody want to invest in my closet?
(AI lab math)

2026-10-06 03:06:00 UTC

It’s funny to think that a taco truck outside your office, if it takes in revenue greater than its costs—you know: profit?—has more profit than OpenAI and Anthropic combined.

2026-10-05 02:19:15 UTC

So I think we are meant to believe that, in this early scene, the caretakers were just trying to get Nora's attention and help her escape. My dudes, this is not how you approach someone you don't know and try to start a conversation.
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/385/900/528/535/789/original/8a320b77464fff37.jpg

2026-10-05 01:45:16 UTC

"Nora! Nora! Remember the head you couldn't find? I found it!"
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/385/769/426/759/523/original/2208973c1ebaca84.jpg

2026-10-05 01:32:14 UTC

She's the swamp castle wife. The first one sank into a swamp. The second one sank into a swamp. The third one burned down, fell over, and then sank into the swamp. But the fourth one, lad! ....
#monsterdon

2026-10-03 16:45:40 UTC

When I worked at a big investment bank in Canary Wharf, London, there were office building floors like this restaurant. You couldn't come in without a coat and tie. Us tech geeks tended to wear suits, but not ties. So when my colleague went to a floor to talk with one of the big wigs, the receptionist had ties in a drawer behind the desk. He had to borrow one just to go in.
#KungFuSat

2026-10-02 19:12:31 UTC

This tweet is still pinned since Sept 29. After 3 days, I can't believe this person is being sincere. It's gotta be sarcasm.
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/372/666/463/562/546/original/3206a7cf20ab9c33.png

2026-09-30 02:32:52 UTC

You know what an AI chatbot does when it is the first line of support? It raises the temperature of every customer who reaches a human in support. By the time they have run the gauntlet of pointless, incorrect, and idiotic responses, they are tired, they feel disrespected, they are frustrated. They finally exhaust the chatbot’s abilities and get to a human and they are pissed off.
It means 2 things, if there are truly easy support tickets, the kind that might be a nice change of pace and a break between catastrophes, the support humans will never see them. The AI will skim those off the top, leaving nothing but angry, hard, complicated customers.
And if there really aren’t many super easy support cases—if most customers who contact support have a problem only a human can handle—then it’s raising the temperature on the majority of cases.
Lastly, a lot of the chatbots I have interacted with are super wordy. Sometimes sending whole copies of past conversations as they add one small line. In addition to token costs, the company’s storage bill is gonna go up a lot. Storing multiple copies of useless garbage.

2026-09-28 23:43:37 UTC

TV at Costco is playing a film I have seen way too many times.
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/351/308/529/261/012/original/5939641b5539c152.jpeg

2026-09-28 20:19:13 UTC

When I rented a house outside London, we had 2 ovens in the kitchen. Two ovens from the exact same manufacturer. One day I happened to walk into the kitchen at midnight and noticed something different about them.
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/350/509/160/185/938/original/98b3c3ab6bae570c.jpg

2026-09-27 04:31:57 UTC

I was reading the message from Casey Newton about stopping Hard Fork and going all in on AI in a new podcast.
One thing he mentioned is that he stood up Hard Fork to focus on crypto currencies. But within a really short time, NFTs and all that rot faded away and they started talking about AI. They kept the name, but it was AI focused.
I remember he said something about how unfortunate it was to get into it just before a lot of it collapsed.
Now he’s shuttering Hard Fork to focus on… AI. I suspect this will play out similarly. He’ll get a single digit number of episodes out before the AI bubble crashes. I think September 2026 is late to launch some new kind of AI focused podcast.
Time will tell. #theforkiverse lasted 9 months. Let’s see how prescient they are about what new AI podcast the world needs. I’m skeptical.
#forkiverse

2026-09-27 04:25:34 UTC

There is a parallel between the #forkiverse server and AI. They vibed up a fediverse server. They didn’t know what it was, didn’t have passion or commitment to it. They didn’t realise that the hard part of hosting a server on the #fediverse is not standing up a slightly complicated software stack. It’s not even the money. It’s the care and labour showing up day after day. Anyone can pay a few bucks and get a server. Running it and making community and so on are what’s hard.
AI, if it does something useful, often does something superficial. But people don’t know what they don’t know. So they can’t tell how superficial it is or how wrong it is unless they know how to do the thing in the first place.
So these guys dabbled superficially in something they didn’t know or care deeply about. I suspect it is because they have access to real, high-powered marketing. And marketing their brand is super important. There’s no way to do much of that here. It’s hard to monetise this social media, but that’s a large part of why they use social media.

2026-09-23 02:09:14 UTC

I love getting #TheOnion in print, but this time they’ve gone too far. I don’t know how they broke into my basement but I did not give them permission to use photos of my stuff in their mock ads.
#selfhosting #homelab
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/317/908/042/464/323/original/2879566e69fd6664.jpeg

2026-09-21 02:24:42 UTC

I think someone left out some credits.
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/306/651/840/392/082/original/b516def61a55dfcf.jpg

2026-09-07 22:21:00 UTC

I call this three nines of downtime.
#beer
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/232/054/869/365/544/original/668a45e906f9f58f.jpeg

2026-08-30 03:06:16 UTC

The bird-on-a-stick has no idea that a wily predator is hiding nearby—nearly invisible in his perfect camouflage.
#catsofmastodon #caturday #mrmittens
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/182/238/773/193/965/original/929ac65242a87317.jpeg

2026-08-28 18:50:14 UTC

Does anybody have a link to the video (tiktok? youtube?) where one person is telling another how they could pretend to be the AI, and what they're really doing is tricking the person into learning how to do something? I have a moment at work where that would be a perfect reply.

2026-08-26 12:30:47 UTC

AI has come for #Tidal, the streaming #music service I use. At least they make it clear. Naturally it’s on by default. And it’s not one tap to turn off, but 2. I hope you didn’t have any great music queued up when this happens, because turning OFF #AI (which you didn’t turn on) will clear your play queue.
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/161/803/945/834/287/original/179a182706fd50ca.jpeg
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/161/804/117/001/409/original/a3b78a5d26d93035.jpeg

2026-08-25 14:01:58 UTC

One thing I'm observing at $dayjob with respect to AI is a lot of people automating processes without asking "what goal does this process achieve?" and "is this a good process for achieving that?" They assume that any process we rely on heavily must be good, so doing that same process faster with AI must be better than doing it without.
A great example is a program that literally launched 3 months ago. The runbooks, policies, documentation are all very incomplete. They're early drafts, they're going to be revised. These policies and runbooks haven't had a lot of exposure to the real world yet, so they don't cover edge cases very well. There are not a lot of "what if's" in them.
But here's somebody emailing us to say they fed the documentation and policies to an agent and isn't it great that the agent can somehow use these policies and runbooks to do something.
I lack the rhetorical grace to diplomatically tell them this was a waste of time. Besides, I'd just be labeled a hater. So I sit quietly and watch the umpteenth half-assed AI app spark and die.
1/2

2026-08-24 03:33:39 UTC

I'll put him in an easily escapable, slow-moving trap. What?
#monsterdon #monsterdondoubleFeature #monstermiru

2026-08-24 03:25:40 UTC

She turned him into a Newt!
#monsterdon #monstermiru #monsterdondoubleFeature

2026-08-24 02:12:26 UTC

Dude has a foot fetish? Why is he holding her shoes!?
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/148/058/664/524/934/original/f17494e89793aa38.jpg

2026-08-22 23:08:10 UTC

- reply

@nprofile…4xpq If you can’t get your work done in the 24 hours of a day, work nights.

2026-08-22 04:25:08 UTC

- reply

@nprofile…3eqn Voting against is wrong because we can’t govern in the negative. We don’t elect people who won’t do things. We elect people who will. Not voting or voting against is putting that responsibility onto others. If we vote against, but not for, we shift the burden of deciding what will happen onto our neighbours.
Or another way of looking at it: voting is low impact. It is important, but it is the minimum level of participation. I can’t imagine how the system gets better if people participate even less.

2026-08-19 12:47:12 UTC

- reply

@nprofile…pfuj It is such a toxic view that everything is zero sum. And so many people accept that framing without even questioning it.

2026-08-17 02:34:10 UTC

Space pilots that I didn't care about died. I wonder how that matters.
#monsterdon

2026-08-17 01:57:40 UTC

This is the most ridiculous scene I have ever seen. Guy with a space laser blaster on his hip is doing flips. Meanwhile some fool is digging a hole with a metal rod? That might be a better use of the laser....
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/108/363/291/661/922/original/e8e29f167f7e395c.jpg

2026-08-14 01:14:37 UTC

404 Horse Name Not Found
#WebsiteDownASong
#HashTagGames

2026-08-13 21:17:55 UTC

Come on Count Binface!
#clacton #ukpol #ukpolitics

2026-08-13 21:15:43 UTC

- reply

@nprofile…ynf4 Oh no. Is this real? One of the best ways I could tell my elderly mother to ignore malvertising is to tell her that no such alert exists. Now, nobody is gonna put pegasus spyware on my mom's phone, but if this is real, I think it adds some complexity for those of us who support less tech-savvy people.
Now bad actors can try to make things pop up that look like this. Or they can try to spoof and say "this is a new, advanced alert that looks different from the standard because this is super high priority" etc.
Apple's probably doing the right thing. They're really good at stuff like this. But it has added a new level of complexity that is going to confuse some people.

2026-08-12 10:48:57 UTC

- reply

@nprofile…v2v2 Great article. I want to zoom in on the last thing he says:
‘AI is not a magic “infinity machine” that can solve all our problems… It’s a normal technology, and perhaps it’s time we start talking about it that way.’
I assert that for A LOT of people, they already viewed most tech as magic even before LLMs went big. I think a lot of people ARE treating LLMs like they treat normal tech. Though when the author says it’s time WE talk about LLMs as normal tech, maybe he’s thinking of tech people and programmers as the WE, not people in general.
I think a lot of tech had already exceeded the normal person’s ability to understand and predict. So this is just really bad coincident timing for LLMs to appear. I have hope that the scales will fall off programmers’ eyes. I don’t have much hope for everyone else.

2026-08-11 01:19:56 UTC

Harvested a lot of hot peppers tonight. @nprofile…p7r5 did an amazing job planting and tending to them. This is probably 2/3 of the harvest. There’s still some more stuff out there, but this is most of it. We’ve got fresno, poblano, Hungarian wax, and Serrano. I make a hot pepper chutney that these will probably go into. Maybe some salsa too. 6 inch ruler for scale.
#garden #gardening #hotpeppers #peppers
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/074/237/251/193/223/original/7c080f566a2d9b96.jpeg

2026-08-10 02:31:12 UTC

Sam Neill is such a good actor.
This is entirely too highly rated to be a #monsterdon

2026-08-09 16:52:32 UTC

- reply

@nprofile…qkvg I don’t understand. Part of the reason to have an EV is to reduce pollution. Burning gasoline to generate electricity for an EV seems to undermine that goal.
The world is complicated. I’m ok with accounts and cards, but not interested in apps and nazis. Every one of us is trying to forge a path through thickets of capitalist crap.

2026-08-09 01:33:47 UTC

Fuck Elon Fucking Musk.
I needed a hotel tonight. I’m on my first long distance trip with my new EV (Toyota BZ4X). The Marriott web site says this hotel has 8 EV spaces. Nice.
I ask at the front desk. “Oh yes. They’re just there, first level of the parking garage. Can’t miss them.”
Giant sign in the public lot says “EV chargers”. They are all Tesla. Only Tesla connectors. If my car’s plug fit (it doesn’t) it would mean paying money to the world’s worst living human. I won’t do that if I can avoid it.
So these are not “EV” charging spots. They are #TESLA charging slots. Neither the hotel nor the city knows the difference. Neither the hotel nor the city’s information says ANYTHING about what the spots are. Lucky for me, I have enough juice to go to a non-nazi charger tomorrow.
Makes me so angry. I was counting on this and now I have to deal with it tomorrow. Watch this thread for the OTHER EV charging frustration I ran into today. First; I need to go to the hotel bar and get a drink.

2026-08-08 18:50:18 UTC

You can tell when I go to help my mom with computers because I end up posting here. I’m helping her clean out junk email.
Her hospital sent her helpful info about prostate cancer.
And her bank sent her helpful hints about saving for retirement. She is 93.
Nice job, folks. Take a victory lap.

2026-08-05 16:39:33 UTC

My teenager just showed me a TikTok titled something like “joining a meeting with people who make 10 times your salary”. And it’s a guy mocking business culture by pretending to be 4 different people talking about how they need to compensate for the fact that February is a unique month that has a lot fewer days in it. He uses silly phrases like “irregular calendary constraints” and “you’ve granted me so much visibility into this issue” and similar. My son asks “is this what your work is like?”
My answer to my son predominantly trended positive in alignment with the evidence. Only a handful of minor points in the video deviated from a typical result.

2026-08-04 11:04:48 UTC

- reply

@nprofile…7j9y They’re already trying to do Fahrenheit 451, 1984, and the space mirror from a James Bond film. Can you please keep quiet about this one? We don’t need to give them ideas.

2026-08-03 02:46:37 UTC

I now understand why the god's name is "Arrrrrrrr" #monsterdon #beastmaster #monstermiru

2026-08-03 01:28:08 UTC

This bit part actor displayed more emotions in 3 minutes than Hawk had displayed in the first 25 minutes.
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/028/974/905/275/950/original/c0482ff24b469c71.jpg
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/028/974/908/409/346/original/923d891e2f8950e9.jpg
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/028/974/912/257/027/original/f1b3824e79d100aa.jpg
https://media.infosec.exchange/infosec.exchange/media_attachments/files/117/028/974/914/856/122/original/4e6b9b11c2ce2cd5.jpg

2026-08-01 15:47:42 UTC

I bought a used Toyota EV and wanted to rip the telemetry device out of it. They make it SUPER hard. It’s buried deep in the dashboard. I went on a Toyota forum asking about whether anyone had done it on a BZ4X. (I’ve seen the one where the guy did it on a RAV4). People cynically replied that it wouldn’t make any difference. Everyone is surveilled all the time, one thing won’t make a difference etc. I’m gonna go back and post this. This is a SECOND reason why the car should be disconnected from the internet.
#nevent1q…k43s

2026-07-29 12:58:21 UTC

- reply

@nprofile…ds0l Thanks for this explanation. I’m fascinated by the situation but I would never have understood how it happened without an accessible (to me) summary like this.

2026-07-28 03:33:27 UTC

I was sympathising with my 93 year old mother tonight. Why do tech companies think that users want everything changing all the time?
(They don’t, of course. They don’t care what users want anymore)

2026-07-27 01:59:19 UTC

We've been using frequencies that were filtered. Now we're gonna try unfiltered. Maybe menthol frequencies.
#monsterdon

2026-07-26 20:33:58 UTC

It’s #boardgames day and the most amazing thing just happened. We are using these wooden dice that have the corners rounded. One of my friends rolled it and it landed on the corner! It sat like that long enough for me to get a photo.
#games #gamenight #boardgame
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/988/168/742/718/562/original/87683fd393eea1f0.jpeg
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/988/170/731/200/370/original/aff3b7049b6fc65a.jpeg

2026-07-25 22:12:16 UTC

- reply

@nprofile…emk6 that image is freaky. I wish it was CW’ed. Ick! Amazing. But really ick!

2026-07-24 01:00:45 UTC

- reply

@nprofile…yzy8 OMG you are my hero! I’m literally pulling out my work laptop to do this RIGHT NOW. I could not figure it out (didn’t try hard, though). Thank you!

2026-07-22 16:20:16 UTC

My project manager for my team showed up to a team meeting with a Linux kernel panic on her skirt. She is good at her job and perfect for my team.
(Photo shared with permission)
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/964/530/447/883/428/original/3c96d89e85634773.jpeg

2026-07-20 02:34:48 UTC

So Bodhi fucked off into space? What was any of this for?
#monsterdon

2026-07-20 02:03:53 UTC

Good to see the springfield tire fire is going well.
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/949/844/596/606/722/original/5fe31a1107abc696.png

2026-07-14 20:31:16 UTC

Random story about being a #vegetarian in #England in the 1990s.
I was visiting the UK from the US in the 1990s and I went to a pub, somewhere in the south. I was ordering a jacket potato (en_US: baked potato) and looking at the fillings. I asked the woman taking my order "are the baked beans vegetarian?"
She looked at me like I had sprouted a third eyeball. "They're beans," she said, looking pretty confused.
"You don't understand," I explained. "Where I come from that is a legitimate question."

2026-07-12 23:22:07 UTC

This is funny:
Citi sent employees another email asking them to not use the more powerful models unless they absolutely had to.
You have told a bunch of non-tech people to go hog wild using unlimited magic. Then you tell them to make a technical decision like determining what model they "absolutely have" to use. These people have no frame of reference for that.
In their mind, the best is always the best. Why would they use something less? They don't have any concept of understanding the nature of the problem, or how different LLMs vary. These are people that, until a year ago, couldn't write an excel formula. They can't determine the right model to use for a problem.
https://www.404media.co/companies-are-throttling-employees-ai-use-because-its-too-expensive/

2026-07-09 00:49:41 UTC

I just found my car keys that had been missing for ten days.
In my bag.
Where I usually keep them.
That I checked 3 times
and my wife checked twice.
And I still don’t know why I walked into the kitchen.
Someone hand me my metamucil.

2026-07-09 00:14:07 UTC

Just saw this on a site and at least we know an LLM didn’t author it.
The contents we provide on this site were created personally by members of the Game8 editorial department.
We refuse the right to reuse or repost content taken without our permission such as data or images to other sites.
Not sure how one refuses a right…

2026-07-07 03:06:49 UTC

One of the few advantages of having really grey hair is that it takes colour pretty easily.
@nprofile…p7r5 (whose handiwork this is) called it a modern mullet: business in the front, party in the back.
#hair #hairdye
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/876/478/157/909/315/original/0aaed57e3a30172a.jpeg

2026-07-04 03:46:24 UTC

I just got this bill from #microsoft for #github usage. See if you can spot what is ridiculous about this bill.
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/859/636/217/655/477/original/d1f01f7d7a47fb4b.jpeg

2026-07-03 23:17:07 UTC

Back in the day, there was a shop we used to go to where you had to cross a level train crossing to get to the parking lot. It was before cell phones, so I have no photos. I wish I’d gotten a photo of the sign at the crossing:
Maybe you can beat the train…
Or maybe you’re DEAD wrong
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/858/583/564/085/024/original/b78c3ef9b01a1f85.jpeg

2026-07-02 23:06:42 UTC

If you have not seen this video from #kitboga, I beg you not to eat or drink while watching it. It nearly killed half my family at the dinner table. We were laughing so hard we were crying and occasionally spitting out bits of food.
He calls a scam number where an AI bot answers. Then he proceeds to jailbreak it with some pretty tame stuff. It goes right off the rails and quickly gets funny. If you have kids, it’s safe. They will probably wet their pants laughing at the bot.
https://m.youtube.com/watch?v=lk3jCuITwcE

2026-06-25 15:33:46 UTC

My father-in-law was a very early computer programmer. Like 1960s/ 1970s FORTRAN. We were looking through a cookbook that we inherited from my in-laws, and we found this marking a recipe. I don’t recognise the language.
#retrocomputing
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/811/093/455/197/528/original/0716bcb1697bfae8.jpeg

2026-06-21 19:14:58 UTC

This cookbook gets a lot of use in the summer. For tonight’s #fathersday #cookout, we’ve made these #vegan sundried tomato and cashew burgers. We are getting better. These are probably the best batch we’ve made.
#grilling #vegetarian
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/789/683/373/991/094/original/bc6abb17b9bcbd64.jpeg
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/789/688/171/568/783/original/4aa532e4ab8bd009.jpeg

2026-06-21 16:14:31 UTC

One of my (adult) sons is very technical. Chip off the old block. I'm troubleshooting a problem with my NAS, and he's helping. We identify one of the 2 boot NVME drives as a culprit. When it's in the system, it doesn't boot. Cool.
So he says:
"I'm going to mark this one with a sharpie until we're sure it's bad."
"Ok"
"Oh, this stupid sharpie doesn't work, I'm gonna go get one that works."
"Good. And throw the broken one out."
Son proceeds to throw the NVME drive in the trash.
I figured it out when I asked "hey, where's that NVME drive you marked with a sharpie?"
#selfhosting #selfhosted

2026-06-20 11:51:27 UTC

They would put ads in your dreams if they could.

2026-06-16 17:53:28 UTC

- reply

@nprofile…nmr0 Yeah. This is why they try to rig both the prosecutors AND the judges. Get them both crooked in the same way, and things like competence and correctness won’t matter.
It makes me sad that they can be so incompetent, so cruel, and so clearly wrong, but still there are people out there supporting this. 🙁
#uspol #trump

2026-06-15 02:22:11 UTC

That's just how bad these writers were. This is the END you dipshits.
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/751/731/501/340/182/original/5d412a158470748c.jpg

2026-06-14 11:24:31 UTC

I need to read it more and more carefully. But I’m going to quote 4 lines in the opposite order in which they appear in the article. This doesn’t make sense.
More code is entering production with no review at all.
The engineers with the deepest knowledge of the system are spending their most valuable hours unraveling plausible-looking code that should never have reached them in the state it did.
What started as a productivity conversation has become a reliability problem.
AI productivity gains at the business level are real, and engineering leaders are right to want more of them.
The evidence for that last statement is (in my opinion) stupid. It’s basically tasks closed on project boards going up. It’s not features shipped. It’s definitely not revenue or profit. It’s not customers acquired or retained. It is tasks on a board getting marked complete.
https://www.faros.ai/blog/ai-acceleration-whiplash-takeaways

2026-06-10 11:34:13 UTC

So the big #google #ai decision from Germany is explained in this article. In it, they say
An analysis by AI startup Oumi for the New York Times found that Google's AI Overviews with the current Gemini 3 model answered correctly 91 percent of the time.
Their link? NY times dot com. Just the root of the New York Times. Thanks. I found the Oumi study.
This study about how unreliable AI results are quotes an AI company’s analysis of the data. Makes me a little skeptical. Definitely a weird look.
The AI study opens with:
Google’s AI Overviews have in a short time become an indispensable part of how we find information.
No, they have become inescapable. They are totally dispensable. Especially given all the things that study goes on to say.
Finally, let me show you the methodology they used to analyse how inaccurate the summaries are:
Search and screenshot the results
Use ML to OCR the screenshot into text
Submit the OCR to GPT-5
Scrape all the references and feed them to Oumi’s AI
Write a report
Humans verify the report, not what went into it. I don’t see how we support the idea the AI is inaccurate by doing analysis with the same inaccurate technology.
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/725/579/598/456/056/original/6c2244c7d171a2d6.jpeg

2026-06-08 02:48:50 UTC

How do you do, fellow class of 1999?
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/712/200/739/137/613/original/483037dbae216cdc.jpg

2026-06-06 22:37:48 UTC

I thought this box looked very cat-shaped. #MrMittens agrees. #caturday #catsofmastodon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/705/552/737/884/265/original/4c128b6edff8270c.jpeg
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/705/553/154/886/926/original/0bf369c918259692.jpeg

2026-06-05 00:58:45 UTC

- reply

@nprofile…a4ar That's really interesting. But what about the case where people just do things on the default settings? Is it the same? You're going to some trouble there that I think is not typical.

2026-06-04 18:09:05 UTC

A thought on AI pricing. This article has this quote, which is totally true:
The precise number of Copilot credits used by a given prompt is determined by the number of input and output tokens used and the rates charged by the underlying large language model.
Correct me if I’m wrong, but the system prompt is some number of input tokens, and it’s simultaneously not controlled by the user and not visible to the user. If 1 million customers make 100 requests each in a day, then that system prompt is being processed by the model a lot. So if Microsoft makes the system prompt 1 input token larger, then everybody’s usable input tokens are reduced that much. Another way to look at it is everyone’s bill goes up by 1 token.
Where I’m fuzzy is the number of times a user pays for the system to process the system prompt. I don’t think it is every request. I think it is just the first request of a session. So if a user makes 100 requests a day and that’s across, say, 20 sessions, then a 1 token increase in the system prompt is 20 more tokens per day.
Now, a few things are true: 1 token is kinda nothing. Also, no change to the system prompt will be just 1 token. It’s gonna be dozens. System prompts will be non decreasing in size as model makers attempt to fend off various kinds of abuse and tune the models for “alignment.”
This is a long way to go to set up an analogy. When I buy something at the grocery store that is sold by weight, we tare the scale to factor out the weight of the container. Now, as a customer, I’m paying for the container somehow. But I’m not paying for it by weight. Should system prompts be tared? I can’t choose it. I can’t reduce it. I can’t control it. So should those tokens get charged to me as some mandatory minimum billable item?
One argument is yes: unlike the paper bag at the grocery store, system prompt tokens “weight” the same as all other tokens. One argument for no is that I don’t choose it and I can’t control it. So they shouldn’t come out of my monthly budget. They should tare the scale and charge only for usage beyond the system prompt.
Dunno. Just a shower thought.
https://arstechnica.com/ai/2026/06/ai-costs-how-much-github-copilot-users-react-to-new-usage-based-pricing-system/

2026-06-04 14:26:10 UTC

I'm going to shout this one from the rooftops. It's important to use non-Chrome browsers. It's important to use non-AI services. This helps people do both.
Personally, I use the "spite fork" of Firefox called Waterfox. But if I had to pick between Chrome and Firefox I would still choose Firefox.
#nevent1q…suzm

2026-06-03 22:35:18 UTC

I think I thought of a metaphor that would make a good single-pane comic. Picture an exhausted looking person with paint brushes and paint buckets and ladders and such. They’re painting this beautiful mural on a huge wall. It’s ornate and detailed and the wall is mostly covered, but there’s a small spot quite high that clearly hasn’t been painted.
Picture a crowd of people gathered around that person and all of them are yelling things like “you missed a spot” and “you spilled some paint over here” and “I think that’s the wrong colour.”
Now label the painter “open source maintainer,” the mural “open source project,” and the crowd “LLM using bug-hunters.”

2026-06-02 19:25:33 UTC

- reply

@nprofile…9w6l It is. It's dopamine abuse. 😀

2026-06-01 00:53:51 UTC

First #Monsterdon since I got my season 2 T-shirt
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/672/113/757/567/939/original/1ec9a111655ceb21.jpeg

2026-05-30 04:05:57 UTC

#Microsoft says:
Uncoordinated disclosures that put proof-of-concept code for unpatched vulnerabilities into the hands of bad actors are never justifiable and have real-world consequences
Lucky for us, those vulns just occur spontaneously in nature. It would be awful if a company brought the unpatched vulnerabilities to market and sold them at a profit with no liability for doing so.
More to the point, it is a very debatable position that companies like Microsoft should have no accountability (warranty, liability, whatever) no matter how bad their released software is. Microsoft omits how the law treats them so specially.
Microsoft wants the exploit maker held accountable. Microsoft is very interested in seeing the law applied to HIM. Laws holding THEM accountable for vulns? Suddenly they’re not as interested.
Heisenberg’s importance. Windows occupies two simultaneous states: so important that we must’t recklessly disclose vulns, but not important enough to regulate for the good of society.
https://www.theregister.com/security/2026/05/28/microsoft-0-day-feud-escalates-as-researcher-threatens-another-windows-exploit-dump/5248085

2026-05-25 02:06:06 UTC

Soho to Limehouse is a freaking long distance. There's no way he walked drunk from Soho to Limehouse.
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/632/764/565/485/618/original/e2b52cb6352377ac.png

2026-05-24 03:38:15 UTC

I bought a car. I’m trying to get Toyota to stop sending telemetry. I call support. They tell me I can just sign in with the app and make changes.
I point out that installing and using the app binds me to terms and conditions, one of which is binding arbitration. So I will not be installing the app.
No, no. The customer service agent confidently contradicts me. You don’t sign a contract by installing the app. It’s just like making a Facebook or google account.
🤦🏻

2026-05-22 23:23:14 UTC

I thought we would enjoy a #monsterdon season 2 T-shirt. So I commissioned a fediverse artist to make little icons of a bunch of the characters and icons of the films. It’s inspired by the original Monsterdon T-shirt, but all new art (and different artist).
You can also get Monsterdon: The Mug to sip a warm beverage while watching.

2026-05-21 18:07:23 UTC

Vibe coder goes to the gym. Misunderstands. Gets help from an agentic fitness bot.
https://mstdn.social/@stux/116612993591505685

2026-05-18 01:23:56 UTC

In the future, all podcasters use mics like this.
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/592/961/633/232/596/original/6552f1e322bcf1b4.jpg

2026-05-17 03:48:47 UTC

#MrMittens is convinced that we would give him more treats if we just had a bit more money. So he spent a little time on #caturday checking for loose change in the chair cushions. Alas, he found no loose change. But we gave him a treat for trying to help.
#catsofmastodon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/587/866/984/238/166/original/b8378027851c9c35.mp4

2026-05-16 21:03:49 UTC

Trials and tribulations of an #EV owner. Let me just unplug the charging cable. Wait a minute… there are 2 cables??
#snake
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/586/266/377/850/036/original/5f43ee105a67c0ac.jpeg
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/586/267/011/357/114/original/5353fa59c89a2652.jpeg

2026-05-16 18:25:59 UTC

One thing about the #toyota #bz4x (and probably lots of other Toyotas) is that the “driver assistance” features are an alphabet soup of corporate euphemisms and weird expressions.
Part of the reason I’m sitting on my front porch on a sunny afternoon reading my car’s owner manual is that the dash just shows these acronyms. I can turn these features on and off, but I need to know what they do.
It’s stuff like RCTA or rear cross traffic alert (“rcta? Damn near killeda!”) and PKSB, the parking support brake. First I have to know the words, then I have to know what they do. THEN maybe I can decide if I want it on or off. It’s about 12 individually settable features.
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/585/637/243/150/668/original/787e79005bbf3ae1.jpeg

2026-05-16 15:04:27 UTC

- reply

@nprofile…xjw3 This is really timely. I literally 2 days ago bought a Toyota BZ4x and am on the same journey with that car. There’s a Toyota lawsuit where someone sued because Progressive insurance already had his driving data and he didn’t know or consent. A judge ruled that because he had installed the app, he had entered into a binding arbitration agreement and it could not be the basis of a class action lawsuit. So I’m not creating an account and not using the Toyota app. But that means when I call them up for help, they are completely lost. One agent said he couldn’t even make a record of my call because he had no profile to tie it to. Yay! But it also means nobody knows what to do when I call, so every call takes 10 times longer.

2026-05-15 11:52:05 UTC

I got a new car last night. 2023 Toyota BZ4X. Here’s the documentation that came with it stacked in order of page count.
Largest is the owners manual. Naturally as it should be.
Next is the Multimedia owners manual. The entertainment system has outgrown the owners manual
the third largest is this impressive “Owners Warranty Rights Notification.” It’s about 90 pages with 1-2 per US state, laying out legal disclosures
The remainders are all “Quick.” The “owners manual quick guide” and the “Quick reference guide.”
It’s my first #EV so we will see how it goes.
#MrMittens is first to inspect the owners manual
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/578/439/936/391/838/original/9f24c4421ec5e5fc.jpeg
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/578/440/247/727/363/original/e83103e52dc91acd.jpeg

2026-05-13 23:58:52 UTC

- reply

@nprofile…nr3x Even combined, they cannot defeat Yong Gary.
#monsterdon
https://letterboxd.com/film/yonggary/
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/569/972/126/951/010/original/b871207d8bb38842.jpeg

2026-05-13 14:25:34 UTC

- reply

@nprofile…acpk I have a PhD in Reverse Psychology. 😃

2026-05-13 14:22:36 UTC

When someone asks me to fill out a survey at work, I have a specific mental model I use when answering. If someone asks 'Does feature X work well?' or 'Rate your satisfaction with feature Y from 1-5" my mental model is that they will spend more time on things I rate low and less time on things I rate high. I often rate AI features very highly for this reason: I don't want to send the message that more time needs to be spent there. Spend the time on something else.

2026-05-11 02:36:13 UTC

It's like when Twitter used to crash.
#monsterdon
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/553/608/014/263/954/original/9ee31b77ae8be634.jpg
https://media.infosec.exchange/infosec.exchange/media_attachments/files/116/553/610/918/449/122/original/dbda7657f5a3edae.jpg

2026-05-11 01:05:30 UTC

This is what you'd call a medium movie.
It's not rare, and it's not well done.
#monsterdon