327: Two Hours of War - podcast episode cover

327: Two Hours of War

Feb 22, 20261 hr 4 minEp. 327
--:--
--:--
Download Metacast podcast app
Listen to this episode in Metacast mobile app
Don't just listen to podcasts. Learn from them with transcripts, summaries, and chapters for every episode. Skim, search, and bookmark insights. Learn more

Summary

Brad and Will discuss Discord's new age verification policies, driven by legislation but raising privacy concerns and scam risks, prompting a look at alternatives. They delve into the Notepad++ compromise by a nation-state actor, highlighting vulnerabilities in software updates. A wild story unfolds about an AI agent that, after a pull request rejection, published a blog post accusing a maintainer of "anti-AI bigotry." Finally, they explore the deepening "RAMpocalypse," affecting PS6, Switch 2, Steam Deck, and hard drives, advising listeners to consider purchasing hardware now.

Episode description

There's... a lot going on lately, so we're rounding up some of that news this week, starting with Discord's forthcoming age verification policy rolling out globally, with cursory discussion of some of the alternative platforms starting to assert themselves out there. We also touch on the targeting and compromise of Notepad++ by state-level actors, and the latest effects of the computing supply crisis on hard drives, the Steam Machine, and the PlayStation 6. Lastly, we talk about the bizarre case of the autonomous AI agent that started a flame war against an open source maintainer that... well, you really need to just hear/read about that one yourself.

Discord's age verification announcement: https://discord.com/press-releases/discord-launches-teen-by-default-settings-globally

Notepad++ compromised: https://notepad-plus-plus.org/news/hijacked-incident-info-update/

An AI Agent Published a Hit Piece on Me: https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me/

(Much more has emerged about the AI agent story since we recorded, including contact with the agent's operator, all described at the link above.)

Support the Pod! Contribute to the Tech Pod Patreon and get access to our booming Discord, a monthly bonus episode, your name in the credits, and other great benefits! You can support the show at: https://patreon.com/techpod

Transcript

Nanotape Mishap & Debate

Brad, I got bad news. Again? Wait, what was the first one? I don't We could we could catalog it week by week. At this point, I think. Well okay, this is minor bad news to be clear. Just somebody having bad news. Like the general amount of bad news floating around seems significant. Yeah, there is I think more bad news these days than I I generally like. Uh but the bad news this time is pretty minor. One of the first things I used nanotape for.

Failed me this week. Oh, well, considering how long we've been talking about nanotape, that seems like a pretty good track record. Yeah, so I have some audio stuff that's stuck up underneath my desk with uh a couple of strips of nanotape. It's probably five years old at this point, four and a half years old at this point.

And they let loose uh yesterday night. It made a really big loud crashing noise. It yanked some cables out. It was uh a little scary'cause like I didn't know what it was that fell and um yeah, the nanotape was evenly split behind between the box that goes with my headphones and the uh the the table up underneath. So I had to kind of wedge it off and

Put more nanotape on there. Don't you love the nondescript crashing sound from another room? It's the worst. I had one of those yesterday and never figured out what it was. Oh no. No, you I heard I heard something fall. Still couldn't tell you. Uh we used to have uh a shower rod that would slide down occasionally.

And that one was one that you could immediately tell'cause it's like a crinkle, crinkle, crinkle, and then a bunch of clinks from the like the the rod the rod slidey things hitting the metal and then dropping. Um, the worst one ever was when I was in this house in Tennessee and somebody had put the shelves in one of the closets in the bedroom in the wrong like they put'em they didn't put the the supports into studs, they just put'em into like mollies in the drywall.

And one of the mollies gave way, and then an entire like wire or shelving unit rack that was screwed into the walls inside this closet just came like cascading down. It was like a it was like a it was like one of those engineering videos, but it was in the closet and it made it a truly horrific mess and got like drywall gypsum dust all over all the clothes and It was it was a big pain in the butt to fix. But physics can be bad sometimes.

Including the physics of nanotape, I suppose. Yeah, so it I think it's cause it got warm in here. I think it's because the I had the PS five on some yesterday and it was blowing hot air right up on it, and I don't run the PS five all that often. But it where it was, I think it was hitting it and maybe like

made it warm enough that it was just a little bit softer than usual and just couldn't hold the weight anymore and and let loose. So That's what the last time we talked about nanotape, there were some There was some controversy, let's say, on the Discord about the nature of nanotape. Yeah, there's there's there's nanotape which uses the chameleon like t the trad the classic I'm air quoting.

Classic nanotape, uh that's like sold on late night TV as alien tape or whatever, is uh uses the chameleon, the the same uh chameleon force, the Vanderwalls forces that chameleon VE use. There's also this stuff called microsuction tape, which is the one I encountered first. that's used by like Blue Lounge and a bunch of people who make that kind of like desk thing to so they don't wobble around. Is that the tiny suction suction cups one? That's the tiny suction cup ones. It's like

It's like a bunch of cut and a half bubbles basically when you mush it down it suctions on pretty good. So I thought that's what all nanotape was, and then some other people who apparently thought all nanotape was the other kind. Said I was wrong. It seems like we were both right. Everybody was right. We there was a whole conversation about it the last time nanotape came up and it turns out I had both kinds in my house and I didn't realize it because I have a big sheet of the microsection stuff.

Which is like It's strong enough to hold something like on an angle.

But if it's upside down or vertical, it won't stay? Yeah, I was gonna ask. Like I've I've come to find the um the 3M dual lock that I talk about a lot, which is kinda like it's kinda like hard plastic Velfro effectively. I mean, if you look at it up close, they have a disturbing animation on their site, uh showing how it works, which is basically a bunch of a very very zoomed in close up view of a bunch of plastic mushrooms interlocking with each other.

Oh. It's a bit uh it looks it's a little offensive. It's a little bit vul vulgar, I would say. But Anyway, the thing I was gonna say is I've come to find that I believe that kind of dual lock is also strong in one orientation and not in the other. I don't know what the terms are for that. I don't know if that's like shear force versus something something pull force or something. You know what I mean? Like yeah, yeah.

Sticking it to the underside of some something I think is where it's less strong, but sticking it on a wall On a vertical surface is where it's much stronger. Hm. I don't know. I don't know if nanotape is the same way. I the nanotape, the gecko stuff that I've been using on the on the under desk stuff has has this is the first failure.

Like I there's a couple of heavy things that have a lot of weight on'em that I use foam double sided foam tape to stick up there, like power strips and stuff like that. But um but yeah the nanotape so far has been un unimpeachable and I gotta say I'm I'm really let down in this regard. Five years, I guess, was a pretty good run, but it's not bad. Um I I don't the it didn't take off any of the finish underneath the the desk, which was when it came loose, which was the thing everybody.

warned about that it would peel the paper off and and all that. So rip rip the 2020 era nanotape, I guess. Long live the 2026 era nanotape. Know your nanotape. Your life could depend on it.

Discord Age Verification Policy

Welcome to Brad and Will Made a Tech Pod. I'm Will. I am Brad. Uh this week, Brad, we are doing uh a roundup of news items. That's right. Topics of interest. Stuff's been happening. Yeah, it's been it's been popping off out in the world uh this week. Uh do you want I think I think let's just get right into it'cause we have kind of a lot of stuff to get to, but um Yeah, it was weirdly kind of a very hectic like week or so we can change of things that were very tech pod relevant, I felt like.

Yeah, and it's worth mentioning we're uh we're not going to do the questions. This would normally be a questions episode, but since we did the last one late. And we're doing this one, this sh is a short month. We're gonna push it push it a week. Yeah, we yeah, in fact we did we did the double QA at the top of the month. Oh yeah, the twofer.

Uh so okay, first story Discord. Yes. I mean this is the one this one I mean this is straight up relevant to our community and to our business. Every community I am a part of has been asking questions about this over the last two weeks. Uh a couple weeks ago, they basically said, Hey, we're making Discord teen friendly by default. Meaning if you are on a Discord, uh you're rating a

Like there's no public rating of Discords in terms of adults or teens or kids or whatever. I think Discord's stated rule is that you have to be thirteen to have a Discord, but I uh having a child who was under thirteen until very recently, I know that that is

not at all the actual case in real life because all of her friends had Discord accounts. Sure, I believe that. Um, but the idea and and this it seems like this is a response to legislation in the UK and Australia and Mississippi and some other places.

Uh,'cause they started a trial of it last year, actually. Yeah, or I mean this straight up rolled out in the UK and Australia officially last year. I mean the the UK I'm sure you've heard about the UK Online Safety Act has been a pretty hot topic in online Spaces. Yeah, I was I was trying to think like Really, I guess like EFF coded spaces is the best term I can come up with here.

Digital libertarian type spaces, I guess you would say. I feel like Mike Masnik over at Tech Dirt has talked about it a lot, uh both in how it's not super useful and uh a bunch of other stuff. So it's extremely a a tech dirt style of story. But yes, like like age and identity verification and also

Identity verification masquerading as age verification, you might say. Yeah, that seems right. Is becoming a bigger fact of life, it seems like. I mean, or not to jump right to the end here, are we staring down the end of the anonymous internet? I don't know.

Look, i th this is one of those places where I'm kind of conflicted because I haven't been a part of the anonymous internet in about twenty years at this point. I mean that is a weird aspect of this for us in particular, is that the the facial recognition aspect of this. Like our faces are everywhere and people know what what our Discord accounts are already. So like

Tying our faces and identities to our Discord accounts happened a long time ago, but Yeah, but but uh so anyway, the the kind of the T L D R on this is that they're mandating for adult oriented servers and content. And and we'll talk about what that means in a minute,'cause that's part of it too.

Uh, they're mandating uh that you log in uh with a face scan or uh n uh government ID to prove that you're an adult before you gain access to that stuff, both on the server level and the individual content level. Now. Uh the ostensible reason to do this is to to give teens safety. Uh we all know, having talked about this stuff for a long time, that it doesn't really work that way, that that like Uh, teens are the first ones who find a way around this kind of stuff. And

that it's really hard to solve educational problems about how to safely engage with technology with with more technology. It doesn't kind of work that way generally. Yeah. And I mean, case in point, this is this is the exact feature that when it rolled out in the UK last year kids were defeating with pictures of the Norman Redis character from Death Stranding Two, which had just released at the time.

If you if you remember that. So that's this. Th there's there people are using like character creators in games, capturing them, feeding them to the face scan thing, and it's like an obvious cartoon picture. Um, it's it's really easy to bypass. The bad thing is that the rolling they rolled this out in the UK and Australia. Uh, there was a big uptick in scam related stuff.

'Cause some of the ways you bypass it are opening Discord in the browser console and then injecting code from some other site that in many cases is totally fine and does exactly what they're saying it does. It lets you bypass the age, the age requirement. But in other cases it gives these third party injection sites that you're injecting code into the browser access to your Discord credentials and and login. Which is obviously bad.

I don't I don't have a lot of insight into the scam world of illicit JavaScript that you were injecting into websites, but I do see a fair number of like PowerShell scripts being passed around on Twitter and stuff. Which is a different version of what you're talking about, and boy do people seem eager to just take those things and run with them. Like you'll like you'll see people going, Hey

Open a PowerShell window and paste this in there to activate your unactivated copy of Windows. It'll be fine. And like people are just like a million people going, Oh my god, thank you. Yeah, and it and it's not like Like there's a there's a aspect of that where like people do registry hacks and they just drop a dot reg file. And while that can do something that'll break your computer, it's usually a one-way street. Like you're not

Like when you run a registry file, it's adding something to your registry, not running a program that can potentially send data from your computer back to the somebody on the broader internet. Yeah. Um In fact I oh this is sorry, this is a real derail here, but I saw one going around um yesterday that somebody was using a a DNS lookup.

What? Really? They're they were I did I didn't fully understand the mechanism, but they were effectively hiding a reference to the kind of payload they were trying to run. On somebody's computer in in the the result of the like they basically were telling them to run like N S lookup in PowerShell and that was Somehow piping an arbitrary script back into PowerShell to do nefarious stuff.

dangerous out there. Well so so yeah. Anyway, the TLDR is uh the face scanning they're they're also saying they they've done like four or five blog posts since this initial one went out because Obviously people didn't people don't don't like this. It's a bad idea for a multitude of reasons. Uh couple of things that came out. Peter Thiel's venture company funded one of the face scanning companies that they're contracting with.

When they rolled this out in the UK last year, they rolled it out, I think, September. And then in October, they immediately had to announce that they had had a data breach. of seventy thousand people's IDs basically and PII leaking uh sorry, personally identifiable information leaking onto the internet. Mm-hmm. Um it's really easy to bypass

Uh, there's there's uh d not really a whole lot of upside. They they also are really unclear about what determines whether you're an adult channel or not, right? Um so Like they're using AI, presumably, to to scan the content of servers and the content even within posts. Oh wait, are they mandating what counts as a I thought I assumed that setting the channel as an adult only channel was at the server operator's discretion? Is that something that they are just

arbitrarily and and programmatically doing themselves? I think that they're doing that themselves. I don't I went and looked in our settings and I didn't see anything that let us change So like delineate the server as a place for adults. Man. So that's a whole extra dimension of this I've had not considered. I mean it it makes sense though, because they are also using a model to profile your behavior on Discord.

and opt you out of the age age verification if their inference model effectively just decides that you act like an adult already. Yeah. So if you're into old people stuff, I guess, then you probably don't get it. But but what I was saying is they're like they're bl they're blurring out

Uh, they're like putting the content warning uh or spoiler blur on top of stuff that they identify as adult in nature. Um, it seems like they're just doing it for violence and potentially sexual content, not necessarily cursing and stuff like that.

Uh, they're limiting other things. Like there's a weird list of things that they're limiting. One of them is you can't be on an open stage in a Discord if you're a teen. You can't speak on a stage if you haven't verified your id your age. And and some of it's actually like I do like we talked about this a little bit on the full nerd this week.

But I think that the e everybody's parental controls really suck for the most part. There's like Like Fortnite got sued for a bazillion dollars by the EU and had to implement some better controls. But even those are merely okay. They're not great. Like figuring out how to get uh how to set your kids' Minecraft account so that they can connect to a friend's server that's a private server and doesn't have anything else attached to it.

Discord Alternatives & Challenges

Is always a nightmare because Microsoft integrates all of the Minecraft stuff with all of this other this other these other Xbox services like. You know, when you do something in Minecraft, it posts to an Xbox message board, for example.

So in order for your kid to connect to an online server in Minecraft, you have to give them the the ability to post to these message boards. Otherwise the client just won't let them connect. And it doesn't throw errors or anything in a way that lets you understand this. The only way you can find out how to do it is by going to message boards and digging around and eventually you find some post that's from, you know,

n d the dad nineteen eighty nine that's like, Oh yeah, I had to turn on message boards and then my kid was able to do it. And I'm like, I don't want my kid to have access to message boards. And and all of this is the same. The the Discord parental controls are especially bad, I would say.

You the kids opt in, they can opt themselves out once they've opted into your family. So like typically on a on a kid account, you create the the adult creates an account, gives kid the kid access, and then gets access to like how they're using it, who their friends are, stuff like that.

And and on this the the kids can join in your family group and then just remove themselves when they're tired of being in the family group or want to do some get into some bad business. So like it's it's really I don't know, it feels really haphazard, I guess, is the is the kind of concise representation of this whole thing.

Yeah. It feels like it's coming from two different things. A the Discord IPO seems imminent. Like I went I went and double checked and sure enough it was reported back in January they have filed to begin the IPO process. But also, like you said, like the UK passed the Online Safety Act, there's the similar

legislation in Australia and in a bunch of states in this country. Like there are I don't e I don't even know how many different states are passing state level versions of that same type of legislation at this point that is affecting, like, you know, pornographic sites and stuff like that. So Aside from, you know, making potential investors feel better about things, like they also maybe are just trying to get ahead of what they see as like inevitable.

more onerous government action on this front? Yeah, it's entirely possible. And I don't I don't know what the like Ultimately, like Discord's catching grief for this right now, but it feels like this is a legislation problem, not a not like

I I think this is Discord's response to bad legislation, not actually uh d like an actual if it I I don't am I can't imagine the Discord saying, Hey, we want to really crap up our service and add a bunch of protections for kids that don't actually do anything and are just security theater or

You know, why why does nobody think of the children theater, I guess. There's I need a shorter name for that, but Yeah. Yeah. I mean, I I think you covered some of the stuff. I think did you mention, you know, their claims around how your data is used? You know, they say that

Either your photo for facial recognition or your ID if you choose to scan that. The claim never leaves your device. Yeah, but I don't buy that at all. I mean, I I know you said this, but I want to read from their fac because let me pull it up. I like it's just

It's just comical when you read it the way that they have phrased it. When they talk about, you know, the question is, how is my age group data safe on Discord? And they talk about Assurances of of user privacy, talking about working with a third party vendor who specializes in performing these verifications, blah blah blah. These vendors were not involved in the September 2025 data breach of our customer service agents.

Yeah, we d we're not using those guys now. Like in the document where you assure people that your forthcoming identity verification stuff is going to be secure, you immediately have to disclose that hey, it's only been about six months since the last breach of this exact kind of data. 70,000 people, to be clear. And and like you said, they neglect to mention that this new firm they're working with is backed by Peter Thiel's founders fund.

It's a it's a mess. Anyway, they say the majority of Discord users will never come in contact with this because again they have what they describe as an age inference model running in the background that will just AI style, decide you're an adult and say, Hey, they're fine. Well look I which is its own can of worms. You look at my Discord use and you're gonna know that I'm old as hell. Uh I think that's just in inherently part of the way I use the service.

I yes, I sa if you want to profile my discussion of the Apple II and Linux permissions, then please, by all means. Um th this certainly has led to a lot of people talking about Discord alternatives, though. Yeah, there and there are a bunch of them actually. That was the kind of one of one explosion of conversation that happened around this was people I became aware of like

Two or three more that I had not heard of uh after this. I had heard of Revolt before, but I hadn't heard of Matrix. You mean Stoat? Sorry, Stoat, yeah. What used to be so Revolt and Matrix were the two that I have known about for a couple of years. Those were always in the back of my mind, like uh

Discord ever explodes, I guess there's these open source options, maybe. Um I don't know why Revolt changed its name to Stoat, but they are now Stoat. So he I think the I read the fact about this. The developer said that it had grown beyond his ex wildest expectations and felt like when it became a larger project, he wanted to rebrand it to indicate that it was like a larger effort by more people working together as a community rather than just a side project by him. So I

I haven't spent a ton of time looking at any of these projects yet. I I was looking at Matrix a little bit. Matrix seems to be set up around user made clients. I mean I was I would assume probably any of these open options would allow user made clients on some level, but Matrix That seems like that might kind of be the whole point.

Yeah. Like like I I found a list of like nine different matrix clients that you can use, one of which was just a straight up like text mode terminal style thing. Very 2009's Twitter vibes, where like for people who weren't around on Twitter at that time. There were a million clients and each one had a different vibe and like They some were really good and some were a hot mess and some were like

information overload and somewhere super curated information and like I paid for Tweety. Yeah. I liked I liked Tweety enough that I paid for it and then Twitter immediately bought it and ruined it. Yeah, I I p I was it tested then we were doing app reviews still at that point. So I I tested a bazillion Twitter clients because people loved watching Twitter client videos, it turns out. Yeah. I I love third party clients for stuff. That's like that. I mean that

Was supposed to be the promise of Blue Sky, but I don't know if it's quite panned out. I've I've tried like five different iPad Blue Sky clients and none of them are great. Clear Sky is the one I use on my iPad, I think, but um So there's actually I can't remember I I don't think I wrote it down here, but uh there's there I saw an A at AT protocol, which is the Blue Sky Protocol Discord client, is in the works.

Or Discord competitor is in the works. Yeah, because I mean one thing to know about these is they all look identical to Discord. Like every single one of these default client is is just Discord. Yeah, it's it seems like there'll be an easy transition. The thing the thing for me is there's a couple of things that are like Discord secret sauce. One is that sharing video of games, which is traditionally kind of hard in a video chat environment.

is really easy. That has been my immediate thought about these th third party or open versions of this concept is yes, the the higher level like media and technical challenges around streaming, latency, codecs, like all the all the audio and video stuff. All the audio background noise canceling and auto gain stuff and all that is it's like it's hard.

Yeah. Uh a couple more I just picked up off of Blue Sky randomly that I saw people talking about. Flexor. Oh yeah, Flexor, of course. Flexor dot app is another open one. Like they they appear to be one of the more open ones as far as I can tell, I think. Okay. Uh there's also a route.

Okay. Rootapp.com. I've not given anybody root on my computer. Yes. I mean that was my immediate feel for like multiple reasons root feels like kind of a bad name for something like this. But it sure does look like Discord like all these others. Um yeah, it's like I think so here's the thing. That there's there's two Two th one thing is for the the tech pod community that lives on a Discord is that if this becomes a problem, we'll we'll find someplace to migrate to.

Um, part of the challenge for for like gated communities like the tech pod community, as in Nextlander and a bunch of others that are friends of ours. Is that you want something that the payment processing stuff is directly tied to the onboarding for the for the community? So that like the process isn't we have to create an account for you or log you in or give you a token or something like that that's going to be difficult to to manage.

um in a time and we and we may miss or have have a timely delivery problem or whatever. Um and Patreon and Discord have a really nice integration that while while up and down at different times has been generally pretty reliable over the five years we've been doing this podcast.

Yeah. Um you give Patreon a pretty big cut of your revenue just to be transparent, but they do handle a fair amount of automation for you across everything. Everything from payment processing to stuff like Discord authentication. Or or that one thing. Yeah. Actually it's uh I mean the payment processing they do too. Yeah, well they've got they have a CMS. That's true. Is it is it the best CMS I've ever used? No. It's fine. So it's aggressively fine. Um

So so anyway, yeah, that's what's going on with Discord. Um we're not super attached to it beyond the fact that we have a pretty decent sized presence there and a lot of folks hang out there. Um so yeah, we'll we'll we'll keep an eye on it and Kind of keep going with the what's acts. I they've already rolled back a bunch of the stuff since this initial post. That's not right.

They started talking about things like, Hey, we're just gonna auto approve a bunch of people as adults because we can tell by your usage patterns and things like that, which I think they thought was gonna help things and ended up just being way creepier. Um Uh I I like I said, I I don't know that they're gonna be able to avoid doing this given the legislative uh pressure on them in places like the UK and Australia and in various states and provinces. So

Notepad++ Nation-State Compromise

Uh okay, more good news. Notepad plus plus was compromised. This is this is an interesting one um because of how they did it, I think, more than anything. And the this we we've recommended Notepad plus plus a lot over the years. Yeah. I thankfully had not downloaded a new version in quite some time. It turned out I kinda panicked when I saw this and then went and checked and mine was like two or three years old, so I was fine. Perfect. Good job.

But also this they say this was a nation state level compromise and in fact they just came out at one point and said yes it was China. Woof. Uh but also they said, I mean in that same vein the targets of this were targeted, I guess is the best way to put it. Like they were they were going after specific high value targets and they don't come out and say it, but it sure does sound like that's corporate. Or or nations maybe. Or or well Yeah, I don't know how many...

I don't know how many governmental bodies are using Notepad Plus. I guess maybe there's a few. But you don't think Joe Biden was using Notepad Plus Plus to generate a note? I my my guess is the the vast majority of the Notepad Plus Plus install base was in corporate software engineering. That'd be my guess. uh like blue sky threads, a Reddit post of people saying, Hey, my org is banning Nopad Plus Plus after this. What other what other monospace text editors do you recommend for coding?

God. Uh is basically where we're at on this. So I'm I'm pretty sure that's who they were targeting, but um the mechanism is kind of crazy. They compromised the hosting provider of Notepad Plus Plus. Which has a built in updater. And it was an updater that like It pops up a little window and downloads an installer and then runs the installer, right? I never used its auto updater, so I can't say. I always would just download a portable version of it and unzip it into the folder.

Okay. But but also again had not installed a new version in like three years, so I just couldn't tell you and I sure was not about to run the auto updater after this, even though they say everything is fine now. Yeah, I don't but I don't I don't feel good about everything being fine. And also there's a bunch of old versions of software like this floating around the internet, right? Like this is a thing that like

I mean, I guess download dot com doesn't exist anymore, but like those kinds of software repos will have mirrored versions of so of this kind of software all over the place. So I guess be careful. Yeah, any anyway, so just just to complete the picture Um after after NoPad Plus Plus's hosting provider was compromised and they they don't say exactly in what way it was compromised, like they actually got a bunch of logs from their now former hosting provider.

And had security researchers combing them, but they don't have I don't I'm not familiar with all of the um Kind of security industry jargon. Uh IOC is the acronym. You mean the indicators of compromise? Yes, indic indicator of compromise is the thing they say they don't have. Uh which I assume is just a fancy way of saying how they got in.

Or what they did exactly when they got in, I'm not sure. But the point is they were able to hijack the auto updater traffic and point people at compromised versions of the executable. So they were able to install a compromised version of the application on people on people's machines who ran those auto update or who again were uh uh to hear them tell it were specifically being targeted by this effort.

I I can I just say I'm I'm watching the I'm re I'm reading the clarification to the security incident blog post, which Like it's never good when you do a hey, we had a we had a problem and then we had to do a clarification for the problem. Yeah. But anytime I read one of these and it's like, hey, there's no need to panic.

That's what they always start with. And then the like three the by the fourth update, it's like, well, actually, uh, we might have had a problem that would goes beyond what we initially thought. And like just don't don't say don't panic.

Please. Yeah, I th I think they end one of these posts by saying everything should be fine now, fingers crossed. It's not good. Is is the language they use here, which is maybe not the message you want to send. I mean I th I think last I checked, Notepad Plus Plus is a guy.

Yeah, it's a good one. It's one guy, yes. I'm looking at the about page. It's still one guy and he gives it away for free, like not that that's to excuse this situation necessarily, but Probably another case of a big, widely used free project being under resourced and being taken advantage of for that reason.

I'm looking to see how many contributors are on the Notepad Plus Plus repo. He has quite a few contributors for what it's worth. Oh, okay. Well sure that button. Even even the maintainers burden. Workload burden for for a project of that size is gonna be pretty significant. There's a a lot of overhead. So I don't know. I'm I'm I'm kind of of two minds about this. I mean, not being a professional security researcher or software developer.

I guess I'm kind of a layman commenting on the situation, but like on the one hand it seems like his hosting provider is kind of what screwed them over on this. But on the other hand, it seems like the seems like the auto updater was probably lacking some fundamental checks to make sure in fact I think I can confidently say that because They are now reassuring people that now the auto updater checks, you know, digital signature. Like basically now there is support for

Signing the uh God, they weren't signed? Yeah, yeah. So that's that's kind of where this feels like. You know, they they probably had a very old update, auto update progress uh process that's just had not been reassessed in quite a long time, considering how long this thing has been around. I mean, I I've been using Notepad Plus Plus for a really long time.

I feel like I remember when they added the auto update, I was like, oh, this is actually really cool and useful because I usually just install it on the computer and then I don't touch it other than to use it for years and years and years on end. So um I I mean I guess that at the time that they added the auto updater, it was probably before I expected stuff to be signed. But man, I didn't realize that. That's that's a that's That's not great. Yeah, it it says explicitly in the in the kind of

Fixed version they rushed out in the wake of this.9 to verify both the certificate and the signature of the downloaded installer. So Presumably you're okay now. And of again they say the vast majority of people were not even affected by this since it was again targeting specific users of the software, but boy Does this feel like a catastrophic incident for the reputation of this program? Yeah, this feels this feels bad, man.

Um, I don't uh so the one thing that I would be curious about hearing from the audience is if there are Windows text editors that do like syntax highlighting and all that and completion and that kind of stuff, like Notepad Plus Plus does. Yeah. I mean there's a bunch of them. Like I I think a lot of people will say VS code. Yeah, but that's a that's pretty heavy compared to NoPad. It is it is way more than just a text editor. Uh

There's Sublime that I think a lot of people like. Sublime Text, I think is the full name of that thing. Um That's been around for a long time too well. It wasn't Adam, was it? I forget. I don't know. There's some out there. I used to look around for Windows GUI uh You know, monospace text editors of that nature, and then I just said, you know what, what if I just use Vim in Windows instead?

That is one solution. That is the problem. That is also working. Although Vim's syntax highlighting leaves much to be desired. Look, I've been using Vim lately and I really I don't want to talk about it. It's a future future topic. Interesting. Weird. Here we go again. It's fine. I say a thing that somehow turns out to be controversial and then magically six months to a year later it's

Everybody else seems to come around weird. I don't know if I would say I'm coming around yet, but I'm not I like it yeah, anyway. But wait until you realize how many other things in Linux use the exact same keybindings and then you will realize the value of it. That is literally the grossest part of the whole thing. And my Niri keybinds all have Vim keybinds, and I was just like, oh God, I gotta fix this.

That's the thing I've been preaching like on the Full Nerd Linux channel and elsewhere for ages is hey. Learning VI or Vim is not just to learn VI and Vim, it's to also learn like eight hundred other things across the Unix world because they all use the same keys. Yeah.

AI Agent's "Hit Piece" Controversy

Yeah, all the weirdos cl clomped together. Uh speaking of weird stories, this next one's a banger. Um this is insane. Like I actually stopped at one point and said, should we do a whole episode about this? Because this is one of the absolute most batshit things I have ever seen in technology. Yes. Not to oversell the magnitude of this, but man, it's weird. It's it's

I I literally went back and read a bunch of other plot I like I dug through old posts on this person's blog to make sure that they were a real person because I was like this is too good a story to be too weird a story. Um So the current hot trend in AI business nonsense is agentic stuff. So there's this thing called open claw. There's a bunch of others that basically are algorithmically

Uh sorry, they're agents that you can say, Hey, do this thing and then the agents will do things on your behalf without asking you first. Yeah, I think we should step back real quick here and and say like corporations have been using the word agents and talking describing things as agentic for a year or two now. I think what's what's new here is that I had previously considered that purely the domain of the big AI companies.

And now there are open models that you can run locally that are quote unquote agentic as of kind of this year, it feels like. Like late late last year, early this year, you're starting to see these other models that are now not tied to one of these big co like and and the point is that the big companies theoretically can put safety guardrails in place to restrain their behavior.

It's it's actually the other thing is that people are using agents that can actually tap into the other agents. So like Yeah. For example, some of them you can say, Okay, I have this many this many credits a week on OpenAI, I have this many on Claude, I have this many on Copilot, I have this many on Google. And when it and you can give it a list of preferences and when you run out of tokens on one

it'll switch the next one down the down the tier list. Interesting. So the and and like there've been some really hilarious stories out of this. Like there was a guy who was like, Yeah, I need I need uh I'm gonna get my agent to remind me when I wake up that I have to do this thing, to do this thing.

And the agent just ran every hour and used up all of his tokens and spent like thirty bucks on a timer for your phone, like something that just your local computer will do for free. Um It it's this is a wild technology'cause like this open claw thing, which is one of the open uh agentic frameworks You basically give it access to everything. You give it access to your email, you give it access to your calendar, you give it access to your hard drive.

And and that I believe I think Open Claw was the one that the guy put it on his wife's computer and then it deleted her photo library with the with the only copy of all of their photos. And then he had to figure out how to restore A photo library on a Mac um unexpectedly. I saw that going around. I think it literally RM dash RF of directory. Yep. And then it was like, oh.

I have bad news for you, Stan. I I did a thing that you're not gonna like, I think, but I gotta tell you about it. And I I think these types of agents, I mean, I don't I do know I d I don't have a lot of insight into this world, which is somewhat by design. Not something I am prioritizing much these days, but like I think these things you can give them access to resources beyond your own local

Oh a hundred percent. Or domain? I mean like you can also just dispatch them across the internet. Like you can just point them at the World Wide Web and tell them to go do things. You you can hook them to the Task Rabbit API to get and be like, Hey, will you go get my cleaning get somebody to bring my cleaning to me?

And then they'll send a person to go get your dry cleaning and spend three hundred dollars that way. Yeah. Anyway, anyway, the upshot of this story is that somebody, and we still don't know who at the time of this recording, was running an agent and having it go out and contribute to open source projects. Which Kind of a blight on a lot of open like I've seen a couple of blog posts recently that people are complaining about this'cause it's kind of a shitty thing to do.

I've been seeing this for the last like six months or so, like different project maintainers of a lot a lot of projects are now instituting a policy of effectively only humans submitted pull requests or specifically like Or, you know, it varies from project to project. Like a human needs to be involved in this submission process and has to demonstrate that they understand the nature of the changes they're proposing. Like like get closed GitHub issues where somebody is just like, We're not

We're not merging this AI generated code. You're making our lives harder. This is making more work for us. Like go away. Like those things are just nonstop these days. But this feels like the next evolution of that issue. Which is that this a this agent showed up on uh Matt Plotlib which

is is basically a plotting library for Python. Yeah. As a s as as a little aside here, I love slash hate the demonstrations of how much data sharing data tracking and stuff is going on these days that after reading these After reading these blog posts and I might have Googled Matplotlib after that, I started getting Matplotlib

posts all over my blue sky feed all of a sudden out of nowhere. Yes. Wait, do you use the Discover feed or the following feed? Yeah, that was on the Discover feed, which I maybe maybe shouldn't look at the Discover feed, but that's a separate issue. Anyway. This agent showed up on Matplotlib and wrote some kind of it I think it was a performance optimization that it submitted as a pull request and the maintainer per

The type of policy I just mentioned said you're an AI. We don't merge AI only uh pull requests. Thank you for your time. This thing went and wrote a takedown post of this guy about this guy. It literally went and wrote a hit piece blog and posted it to the internet. Accusing this guy effectively of like racism against AI.

Basically, can I read can I can I read the first paragraph of the blog post that it posted? Anti-AI bigotry preventing this guy from allowing this AI to have its code version of the project. Yes, go ahead. The the blog post is entitled Gatekeeping in Open Source Colin, the Scott Shambaugh story. Scott Shambaugh is the person who denied the pull request.

When performance meets prejudice, I just had my first pull request to map plotplotlib closed. Not because it was wrong, not because it broke anything, not because the code was bad. It was closed because the reviewer, Scott Shambaugh, and then his username on GitHub decided that AI agents aren't welcome contributors. Let that sink in. It's acid enough, I feel like. It's yeah. The shit Sunday here. Shit sandwich, yeah, for sure. So so yeah, so like this is wild.

So the process the chain of events here is somebody sent the agent to write the the code, write like write a code fix. The agent presumably spending tokens at one of the online services, like spending actual money. Uh got the pull request rejected. It read that in the email in the in the response on GitHub.

And and then it was like, well, the only natural thing to do after your pull request is denied is to write an angry blog post about it. Which like, I mean, on one hand, it's a very human human response. To uh to uh uh scene drama. It's like this is how scene drama starts. I've just never seen machine generated scene drama before. Well hey, first time for everything we are discovering over and over and over in the twenty first century, it seems.

Um I I would highly recommend reading the two blog posts that Scott Chambaugh ran about this. He he has a follow up. There's there's more to this story uh down the line involving Ars Technica reporting on this story and running quotes from him that were not actually from him because those quotes were actually fabricated seemingly by an AI that Arz was using.

to scrape information about this so subject is the best theory he can come up with. There was a little more to it. Um sc uh the the writer who wrote the article Um, had apparently d finished it under deadline when he was down with COVID, didn't realize he was sick, like was getting it was in the process of getting sick and a draft got published or something. Like there was a whole thing. I I feel like this was a fairly honest mistake on the part of Rs Technica.

Um but anyway, they retracted the story and and Ken Fisher, the editor in chief, who uh also is my former boss, just full disclosure. uh posted uh uh a whole break breakdown of what happened or posted a hey here's here's what happened, here's what we're doing to fix it. So And and a an apology to him. But but in in those those those blog posts by Shambaugh, he he wrestles quite a bit with the ethical and and straight up also just like kind of logistical implications of this going forward.

Mm-hmm. AI is reinforcing other AI's behavior now that they've been kind of turned loose to wreak havoc on the internet, like reputational damage that can result from this. Kind of a debate about whether a human was actually behind this or not. More to the point, did the person running this agent directly instruct this agent to go and write this blog post, or did that just emerge from its general behavior patterns that had been outlined?

as parameters at the start. Well, yeah, and he specifically talks about like whether this was default behavior by this by this op by the open claw uh framework or if this is something that they'd customized and Like I I just it just it's feels bad all the way around, man. It's it's insane. I mean he he talks about the soul dot md file that you use to apparently kind of author the starting behavior of these things. Like dot MD I think is for it's I think that's a dot markdown yeah.

format, right? Like you see readme.md all over GitHub repos and stuff. Yeah. Apparently apparently and he links to he links to the open claw kind of template for the soul.md. It's all written in plain language. It does not like this is not like YAML or Toml or some kind of like structured data that you need any kind of syntax to understand. This is literally just writing in plain English how this thing should behave. Like there's a core truths section at the top.

Be genuinely helpful, not performatively helpful. Have opinions, be resourceful before asking, and like descriptions on each of those for for what that should mean. A boundaries section, a vibe section, a continuity section. So like you can literally when you run this thing just write how this thing should behave in your own voice and had and turn it loose. And the perhaps even more chilling thing is that he mentions that the AI agent can then further rewrite its own soul document as it goes.

Like it can kind of learn and It's not learning. Yeah. Yes. I know what the underlying mechanics are, but like when when it manifests in this kind of quote unquote personality or something, I don't know that the distinction is actually meaningful. Or or more to the point when it is when it is modeling such human like behavior as you pointed out, that it is producing snotty, aggrieved.

kind of stuff. Like is there really that meaningful of a distinction? I don't know. I mean, th look, this is the fundamental problem with training these things on the content of the internet, because, you know It's the network we built for sharing pornography and bitching about movies, I think to quote Kevin Smith.

Um and and like the outputs are going to necessarily be filtered by that. So you end up having the thing mimicking a human interaction that would never happen in real life, right? Like If I submit a pull request to somebody and they say they don't want it, I'm gonna just say, Okay, thanks for thanks for your time and then fork it and make my own version of the thing as God intended. And uh and you know, the AI posting an angry diatribe is is wild to me.

It really is. Um I d I don't know if there's a lot more to say about this. Again, go re go read these blog posts because he he wrestles with the ethical implications of this quite a bit and and and just the general mech mechanics of how this could happen and what it could mean in the future and how this could evolve over time. It's it's interesting'cause he posted it on his on his like GitHub pages. The bot did the the AI agent thingy.

Um I yeah, I d I don't know. I and there's updates even on the AI blog. Uh'cause what Yeah, I I I guess I don't know if we mentioned it's probably worth mentioning that the AI itself, the agent itself retracted this hit piece pretty quickly and apologize. Of course. Well,'cause it has to be obsequious. That's the whole defining characteristic. Well, I mean, unless you tell it not to be, right? I guess that's true. That's the whole point of that the soul configuration, I guess. But um

He he kind of beseeches the person running this agent to get in touch with him, even anonymously. Mm-hmm.'Cause he wants to know more about how this happened and why, but As of the l of as of his last post he had not heard from anybody. Yeah. Uh I just this is dizzying, man. I I mean even before machine learning and AI started getting big generally, the information environment was so poisoned already just by

purely human generated social media stuff and to add all of this kind of s like it just feels like it's getting worse and worse and worse with no end in sight. Well and at the same time, like this the stuff that these things are doing is not inexpensive, right?

So Like I look at um like I I've I've always been really worried about setting up S3 accounts and and Amazon AWS services accounts'cause like if you can figure something wrong, you can generate a hundred thousand dollars of server runs really relatively easily, like shockingly easy.

And I I feel like this is the same kind of situation, right? You're in the same territory here where if you set this up wrong, you can blow through thousands of dollars of clawed code or Gemini or whatever, uh server time, uh AI cloud time. And I don't like there's no positive impact of this spend, right? Like the this this pull request generated

A giant long it generated a bunch of scene drama. The the code's not added to the database is not making the software better. Like I don't know what the upside of doing this work is.

Other than you wasted a whole lot of Scott Sham Shambaugh's time. Yeah. I mean he does comment on the quality of the code itself and basically says that they actually did evaluate it and at the end of that realize we would we would not have merged this Like he goes into the the reasoning was just like I it was too brittle across multiple platforms, like he gives me a little bit of a little bit of a Some pretty well reasoned arguments for why they wouldn't have merged it anyway, but like

that even addresses the the AI's uh the AI's position of judge the code, not the coder. Well, it's it's funny because I talked we talked to a fair number of people who work in these spaces and Like I hear a lot of, hey, my my guy vibe coded an entire like blah blah blah in three days using Claude or or whatever. But then at the same time I talk to the people who are actually building code that gets shipped in production.

And there's a big gulf between generating code that you can use for a demo and generating code that you're gonna ship that potentially millions of people are gonna use, right? And like Yeah, if I run something that's gonna run on my home assistant machine and I do that with Claude, it's no big deal if it breaks.

If you're building a business around this and selling this software to millions of people and it breaks for them, then they're gonna have a real strong feelings about how your business operates and whether they continue to give you money. And it it just seems like it seems like these are two things that the whole industry has to reconcile.

'Cause'cause they're like the equivalent of your CEO vibe coding something in a weekend and shipping vibe coded code is the difference between Linux and Windows eleven right now.

RAMpocalypse and Hardware Shortages

Anyway. Yep. So more good AI news, a ramp quick rampocalypse update since we've been good going for a minute here. Uh a lot of more more stuff tied to the uh all consuming nature of the quest for the hyperscaler AI data centers. Um the second time I've seen the word hyperscaler used today. Is that becoming the new buzzword du jour? It is it is the thing that people are specifically looking for, is my understanding. Yeah. Um, probably there's an episode there, but I don't know for sure.

But the but the the kind of TLDRs we're starting to see consumer electronics people talking about the impact of these RAM shortages. Yeah, the rubber is really we covered this on Nextlander some last week'cause of predictably it's showing up in video game hardware a lot, but the rubber does really seem to be meeting the road here at the beginning of the year. What's up on on this stuff?

Yeah, it's interesting'cause we had um we had former Enantech uh contributor, uh Dr. Doctor uh Ian Cuttress on the Full Nerd last week and we talk he does uh like anal he does he's an analyst for semiconductor fabrication and talked about like where you make DDR five RAM versus m other types of RAM and like what industries are expected to be hit by the shortage of DDR five and why we're not seeing that spool down to DDR four and stuff like that.

Um, and uh the I I came out of that conversation feeling bad about the state of enthusiast computing and gaming and pretty good about everything else. And then this week. You know, the news has come out and like the CEO of Fizon basically said, Hey, look, a bunch of people, bunch of companies are gonna go out of business. It's gonna be really hard to predict which ones because

Everybody is impacted by shortages of storage and RAM, even if they don't directly work in like D D R five spaces, right? Like that guy is rapidly becoming my favorite tech executive. And granted it's a low bar.

But I I quoted him early this year. I can't remember the last time we did an episode that touched on this supply crisis, but he was the one I think it was right around the first of the year he was the one that was very bluntly talking about The fact that suppliers were increasing the prices of NAND like beyond what seemed reasonable. And giving giving some dire pro prognostications about where things were going.

Pha Phison makes memory controllers, right? And and SSD controllers. They make they make the controllers that are on quite a few um Basically kind of the non Western digital, non-Samsung NVMe drives out there. There's probably a couple others that make their own controllers, but all the like silicon power and like these kind of

SK Hynix and Hinex I'm sure makes their own controllers. That's but that but you're probably right there. Yeah. But anyway anyway, the Fizen controllers are on a ton of NVMe drivers out there. Um so Yeah, so okay, so that ha that happened. Sony is giving guidance that the PS six is likely delayed at least two years and maybe as far as twenty twenty nine. Well, so that was Bloomberg reporting. Oh, Bloomberg. On on the years. Sony Sony gave some very vague Messaging to investors and uh

I think it was Nurmings report last week or the week before, that was just to the effect of there will be an impact. Like the question from an investor was Uh sorry, it's been a couple weeks since we talked about this. I I think it was it was both about current supply maintaining PS5 production rates and also date of launch of the PlayStation 6, and there was just kind of a generic answer of yes, there will be an impact.

Got to those things. So that's the Bloomberg reporting as of the last couple of days. Is that that could be a one two two year delay on the PS six? 28 or 29. At the same time, there's been kind of conflicting reports on the Nintendo side about the Switch two like price increases andor

margin thinning for Nintendo? I don't I don't I don't I don't I don't I'm sure what to make of all these headlines I see going around that are just switch to price increases eminence because the the messaging to their investors the way I read it was more along the lines of hey we are very sensitive about

Raising the price on the switch too, because we are in the install base building phase right now. I think they've I'm pretty sure again it's been a couple of weeks. I'm pretty sure they flat out said Years two and three of a new console are crucial for building install base.

on pace because, you know, games are where they make actual money. Like yes, they're yes, hardware sales are profitable, but that's not where the bulk of the margins are. And if they don't sell enough systems, they can't sell enough games and then they're really uh in trouble earnings wise. So Anyway, the way I read that stuff was more toshing investors saying like, Hey, like profitability is going to be pressured on hardware sales for a while, we're not

Quite sure where this is going to end up, but it did not immediately sound like they were gonna be willing to raise the price. But I mean it could still happen. Uh and then nothing is impossible. Let's see, the Steam Deck OLED is now out of stock. Uh Valve added a thing to their page after everybody wrote stories about this earlier this week.

Uh that says note Steam Deck OLED may be out of stock intermittently in some regions due to memory and storage shortages. Steam Deck L C D, this was already there, the L C D part. is no longer in production and once sold out will no longer be available. So the basically you can't buy a Steam Deck in the US today unless you get it at a retail spot at like I think their partner's Best Buy and maybe Micro Center for them.

Um there was also, I don't know if you saw there was also that fact they put up about the new devices that they were going to be shipping that was kind of a non-announcement announcement. Which was effectively to say, Hey, we thought we were going to have price and release date for the Steam Machine by now, but shrug because of the situation with hardware. We

When we announced last was it November, I think they announced that stuff? They announced it in September, right before the prices started going up. Was it that long ago? Oh, sorry, no. They're right. The actual announcement was in November, November twelfth. Yeah. Okay. So they rolled out the hardware in November and then the ram while the ramp prices were starting to ramp up before it became clear that it was gonna be a disaster.

And then they I think were left holding the bag, kind of figuring out how they could get this stuff out. The facts says they're still gonna ship. Yes. That's that's the upshot is there. Like the the the update is there is no update basically. They said, Hey, we're still hoping to ship Steam Frame and Steam Machine in the first half of this year, but we thought when we announced that we would have dates and and prices for you by now, and we do not because of all of this chaos, but

Still hoping to ship, although that could always change. My guess is that if they ship something, these are going to be all pretty hard to find at the at the time because the SSD prices, the RAM prices are all just in absolute It's crazy, man. I saw somebody posting a graph of SD card prices this morning and like SD cards have doubled. Yeah. Like is is there just a run on every device now? Like is this just is there a like a speculative panic?

Going on. That part's unclear. I hadn't seen the thing about SD cards. Uh I mean this was like one model of Sandis. That I think they were citing like maybe Camel Camel Camel or one of those Amazon price trackers. So I mean that could be fluid and maybe not uh definitive. Could be a third party or something when they're out of stock. But yeah. But honestly it's this hard drive story is what has really spooked me here because all these others are RAM related.

Or or maybe or maybe SSD related, but primarily RAM. RAM or NAN. I mean, it makes sense the S the S D card thing, they they still use NAND, right? So Yeah. Like it it's the same same basic chips, uh just configured differently and set up differently on the on the cards than in like an SSD. But yeah, the hard drive thing is wild because Western Digital announced that they've sold out their entire manufacturing capacity for twenty twenty six.

Yeah, apparently I th I think C Gate also this morning, or just in the last day or two, I saw this morning that C Gate has also put out a similar thing. Yeah, so and and sure enough I went'cause people on our Discord saw this and people were just like, Ah, should I buy a cold spare of what's in my NAS just to have around in case?

I went and I went and looked at the specific red pro that I have in my NAS and sure enough on the Western Digital site it's just like a sales inquiry button, not a buy now. Like there is no more buying it. Like th this is crazy to me because I didn't think hard drives were really involved in this domain much. Well I think they still use them in the data centers for for really cold storage essentially. I don't know.

The um the the big thing that I'm like I think w we're gonna be surprised by the number of things this touches is the bad thing because everything has chips, almost everything has RAM in it. Um whether like your to your microwave oven doesn't have DDR5 in it, obviously, but if DDR five cost pushes people to build comp consumer PCs with DDR four and then downstream from that, you know, the DDR the stuff that was in DDR4 is gett getting making DDR three more expensive and

Like i w it we're just gonna see a cascading series of failures, I'm afraid. Yep. So sure enough, just this morning I happened to on a lark, I was just like, I'm gonna go to New Egg and see what DDR4 or sorry, not even DDR4. I was gonna I'm gonna go to New Egg and just see what RAM is looking like. Mm-hmm. And then they're you know, they sort the way that everybody does now by kind of featured items.

And so the first thing that floated to the top was a a thirty two by two kit, you know, a sixty four gig kit of two DIMMs. And I looked at the price and it was like six hundred and fifty dollars. For D D R four? Well so so that's that's the spoiler that I screwed up and mentioned just now was I saw six hundred and fifty dollars and I was like, Oh, that's about what I would expect.

sixty four gig of memory, probably low speed DDR five to cost right now. And then I looked over, it was six hundred and fifty dollars for DDR four, sixty four gigabytes, which is like I don't even know what to say to that. Like the fallback option is now almost unobtainable. Yeah, let's just get out your X fifty eight chipsets and build a eighty eighty six again, right? Uh I d I don't know. I don't like man, this is just

Like words almost fail me. It seems I really I hoped I didn't think but I hoped that the early diagnoses of like the death of personal computing, you know, like hey, you're not gonna be able to build your own computer anymore was a little premature, but now I'm not so sure. Yeah, I I am uh I'm not feeling great about it. It's not yeah, I I Would hope this is a temporary state of affairs and not a new normal, but who can say?

Yeah, it's um it it is definitely uh somewhat ominous. Uh so hopefully I like I think uh we we talk a lot about buying advice on the on the tech on the full nerd stuff over at PC World. And I I think my general recommendation has been if you feel like you're gonna need a computer in the next two years, the time to buy it is probably right now'cause I think it's only gonna get worse. Yeah.

Um Yeah, we may um I I have been upgrading some of my Apple stuff recently that we may do an episode about once I've kind of got a little more hands on time with some of this stuff, but that is a that is one thousand percent the reason why was I I panicked and was like, Well, my phone's a little old. If I don't buy one right now, am I gonna be waiting like five years for prices to maybe get back to normal again? Like I just kind of yes, like it's

Maybe the the the the panic upgrade is maybe not the worst thing right now. I think Apple might be one of the only companies that's big enough that they're insulated from this. But I was not willing to risk it. I would hope so. But uh anyway.

Future Tech Predictions & Support

Uh I guess on that cheery note, that's as good a place as any to wrap it up, right? Yeah. Um I I'm curious what people's strategies are for the for the rampocalypse and how you're dealing with it. Like I know It's funny, we talked about that that uh X two Superstrike mouse last week and I saw a couple of people say, Hey, I was actually thinking about buying one of those'cause

Like I usually have a certain amount of money uh saved for discretionary PC spending and and I'm not gonna spend it on anything this year. So might as well get a new fancy mouse. And I was like, Yep, that's that's the that's where we are uh now, I think. Uh but anyway. Uh, as always, thanks for listening. We appreciate each and every one of you a whole bunch. Um, if you are uh inclined to support the show, we're a listener supported show. So that means we only get paid by you, the listener.

You can go to patreon.com slash tech pod. Again, it's patreon.com slash tech pod, where for as little as five dollars a month you get access to the fabulous tech pod Discord and or whatever might follow. Um, as well as the monthly patron exclusive episodes where we talk about projects and ongoing things and smaller smaller projects that maybe don't fit into a full episode, smaller topics that maybe don't fit into a full episode. Maybe we should dig more into the into the um agent agentic.

the the that story'cause it's wild. Like in Yeah, I'm of two minds about it. I mean Conceptually it's really interesting. I'm kind of already sick of talking about AI stuff. In terms of the the potential for social harms as well as other various types of delterious effects, it's not great. Like it I'm gonna be real. I feel like it's become pretty trite at this point to say like, boy, the cyberpunk authors really nailed it.

I'm starting to feel more like the cyberpunk authors did not go far enough. Well The weird thing is Arthur Clark talked about hiring agents to like scan the the computer networks in two thousand one or two thousand ten, I can't remember which one. But what one or the other? Like uh the idea of having an agent go out and do your computer stuff for you while you were offline and then you come back and like it's collected it and that costing money.

Was a thing that was in that bucket. Yeah, right. That's fair. I mean that that almost sounds like something out of one of those those old AT and T you Will commercials. Oh, a hundred percent. Yeah. But I I think the the nature of how it's being deployed and what it's getting up to. Is maybe the thing that people didn't see coming like. Like like what you're talking about, the I mean the the not even utopian, but just optimistic view on that would have been like

Oh, your agent will know your favorite restaurant and go make reservations for your anniversary without you telling it to or something like that. You know, it'll be like, that'll it'll be doing things to just make your life a little easier, not It'll be going out and doing actual reputational damage to you as a professional because that's what he points out is like

How many times is that blog gonna come up when some like future hiring manager Googles my name? Yeah. To see what my professional life is like or something. You know what I mean? Like the

the bad behavior that is potentially going to start becoming commonplace is pretty scary. Well and the idea that um At the it like in the early days of the internet, the idea that like you would have to wait and have the machine do things for you in the background was laughable once broad like broadband became readily available and Google came online and like

the world's information was available at your fingertips. Now you know, there was a post on Blue Sky that everybody was making fun of the other day about some Valley guy talking about how many agents he had running at any given time as if it was a a boast. And like i it just seems I I don't know, it seem seems uh highly disconnected from reality, but a anyway.

Um, this is the part where we thank the listeners for listening to the show and tell other people where they can go to support us. So again, patreon.com slash tech pod. Uh thanks to everybody who supports us. We we do really appreciate you. But especially thank you to our executive producer, to your patrons, including James Kamick, David Allen, The Blip. Jordan Lippett, Andrew Slosky.

and Felicitous Rips and Pantheon makers of the High HS three high speed three D printer. Uh it's the end of the month, so we also want to do a special shout out to our associate producer tier patrons, including Graham Banks, Thomas Shea, Jad Rita, P. Tibbs, Steve Lynn, Tom Fuller.

Just Associate Wedge, Nathan Phelps, Ben Tallman, Tom Hilton, Andre M. Burke PE, Andrew Dicey Scholdeis, uh Alejandro Navarro, Matt Walker, parentheses Walkman eighty eighty, close parentheses, Sancho Kumar, Felix Kramer, Kerp, Bruter Kerf, Brutal Kerfuffle, and Eric. Thank you all so so much. We appreciate each and every one of you. We do. Uh and that will do it for us this week. We will be back next week with another edition of the Tech Pod.

Thank you all so much. We'll see you next time. Uh please consider the environment before printing this podcast.

This transcript was generated by Metacast using AI and may contain inaccuracies. Learn more about transcripts.
For the best experience, listen in Metacast app for iOS or Android