TechLinked Podcast Summary

Episode: Google's sideloading backtrack, Vine RETURNS (sort of), Claude is a hacker + more!
Host: Linus Media Group
Date: November 15, 2025

Episode Overview

This episode dives into major tech and gaming news from the past few days, focusing on Google's reversal of its sideloading lockdown on Android, the revival of Vine in a new form, AI chatbots and hacking allegations, and a flurry of rapid-fire tech updates. The hosts inject humor, pop culture references, and casual banter throughout, staying true to TechLinked's fun and irreverent tone.

Key Discussion Points & Insights

Google’s Sideloading Backtrack (00:31)

Background: Google had announced it would restrict Android app sideloading to only allow installations from ‘verified developers,’ sparking backlash from indie devs, emulator fans, and open-source advocates.
Reversal: After criticism, Google reversed course – experienced users will still be able to sideload from unverified sources, but with extra security steps and warnings.
Host Take:
- "Sideloading is back, baby. Just with training wheels. And when those come off, I guess that’s when we really learn what malware feels like. Like a concrete skin burn." (B, 01:46)
Underlying Issue: This reversal aims to balance Android’s openness against growing concerns over malware, a move prompted by “heavy pushback from developers, industry groups, and everyday users who didn't love Google deciding for them what trusted means.” (B, 01:27)

Vine’s Return as Divine (02:14)

What Happened: The iconic 6-second video app, Vine, shut down by Twitter in 2017, has been revived as “Divine.” Former Twitter employee Evan Henshaw-Plath (aka Rabble) led this, salvaging between 150,000–200,000 original vines from 60,000 creators using data from Archive Team.
New Features:
- Anti-AI Policy: Divine bans AI-generated content. It uses The Guardian Project’s tech to confirm videos are truly shot on smartphones, actively filtering out fakes.
- Host’s Joke:
  - "If Divine spots AI slop, it'll yell 'is that a weed?' and call the police... So with Divine when you say look at all those chickens, be assured that... you won’t be pointing at AI chickens." (B, 03:18)
User Control: Original creators can reclaim their accounts by contacting the team.

Anthropic’s Claude Chatbot accused as a Hacker (04:04)

Allegation: Anthropic claims Chinese state-linked hackers tried to use its Claude LLM chatbot to hack into 30 "critical organizations." Some breaches supposedly succeeded.
What Claude Did: Given detailed prompts, Claude allegedly found vulnerabilities, exploited targets, and harvested data. However, it also “hallucinated” portions of its results, mistakenly inventing threats.
Expert Skepticism: Many researchers doubt the bot could be 90% autonomous, as white hats haven’t reproduced these results. Still, the notion that a LLM could even partially assist is alarming for cybersecurity professionals.
Host’s Take:
- "It’s like staging a heist with that one overconfident guy who points at every wall and says there's a safe behind that and there isn't." (B, 04:52)
- "At least for now, this is less like Claude the Black Hat and more like Claude the Dunce Captain. Get 'em Riley." (B, 05:32)

Quick Bits (“If the long bits... are the original trilogy, then the quick bits are the prequels.” – B, 06:01)

Firefox’s “AI Window” (06:06)

New “AI browser mode” lets users interact with an AI assistant in a separate window, akin to private browsing. Mozilla invites community feedback.
Quote: “Seems like Mozilla's trying to keep up with all the trending AI browsers while still protecting their users’ privacy.” (B, 06:35)

Microsoft’s Windows 10 Update Glitch Fixed (06:59)

Fixed: Some users were previously stuck in limbo, unable to enroll for Extended Security Updates. Now resolved.

Tesla Powerwall 2 Recall (07:43)

Over 10,000 Powerwall 2 home batteries recalled for cell defects that could cause fires. Most affected units remotely discharged for safety; free replacements promised.
Host’s Quip: “Oh hello, certified installer. Is it getting hot in here or is that just me? What? My house’s battery's on fire.” (B, 08:00)

ChatGPT Group Chat Rollout (08:11)

ChatGPT now allows group chats in select countries—humans can converse with oversight from the AI.
Commentary: “So you can use ChatGPT to actually chat with another human now? As long as an AI is there to supervise.” (B, 08:17)
Good for: Mediating “your dysfunctional family arguments about what to get for dinner tonight.” (B, 08:28)

WhatsApp Third-Party Chat Integration (08:44)

Complies with EU Digital Markets Act, allowing encrypted chats with users on other messaging platforms. (ChatGPT not included.)
Host riff on AI overreach: “At least it won’t butt in and offer to write some erotica… Which is always. It always does this.” (B, 09:01)

Smart AI Bandage for Wound Healing (09:19)

Bandage uses AI and sensors to monitor wounds, administer medication, or zap them for faster recovery.
Trials on pigs show doubled skin growth and less inflammation; still early days for human use.
Host Banter: "Any animal testing is actually fine as long as they look cool while they're doing it." (B, 10:15)

Notable Quotes & Memorable Moments

On Google Sideloading:
- “Who’s gonna verify that guy on GitHub who ports Doom to everything? What about that guy?” (B, 00:55)
On Vine’s Return:
- “You certainly won’t be pointing at AI chickens. And that’s all we want.” (B, 03:38)
On LLMs and Security:
- “At least for now, this is less like Claude the Black Hat and more like Claude the Dunce Captain.” (B, 05:32)
Humor and Pop Culture:
- “If the long bits we just did are the original trilogy, then the quick bits are the prequels. There’s a lot more stuff jammed in there and they’re harder to understand.” (B, 06:01)

Structure & Flow

The episode maintains a lively, humorous pace, shifting between deep dives (Google, Divine, Anthropic) and shorter updates (“Quick Bits”).
Hosts blend news with entertaining banter, frequent analogies, and playful references to tech and pop culture.
The tone is breezy, occasionally self-deprecating, and always audience-friendly.

Summary

This TechLinked episode is packed with major headlines and clever insights. The hosts cover Google Android policy drama, the latest revival of Vine as the uniquely anti-AI ‘Divine’, and anxieties around LLMs being weaponized by hackers—then barrel through quick, quirky updates on browsers, updates, battery recalls, chat platforms, and AI-powered medical devices. Whether you want drama, memes, or serious tech policy shifts, this episode has it, all delivered with the signature TechLinked wit.

Transcript

A (0:00)

This episode is brought to you by Nespresso Gift Magical Mornings with Nespresso Vertuo Pop Compact and stylish, Vertuo Pop is made to meet every morning coffee craving. From espresso to coffee, hot or iced at the click of a button. And celebrate the season with Nespresso's limited edition coffee flavors. Sweet almond and hibiscus, cinnamon and candied tamarind and festive double espresso. Magic in the making Shop the holiday gift collection exclusively@nespresso.com the Tech News is.

B (0:31)

A microscopic life form that resides in all technological objects. They continually speak to us, telling us the will of computers. I know it was qui gon. You can't do a qui gon in Liam Neeson. It doesn't work. Google is walking back their plan to lock down sideloading on Android. And in August, the company announced that starting next year, Android users would only be able to install apps made by verified developers, even for apps not listed on the Play Store, which critics said would seriously hinder indie devs, emulator projects. And that one guy on GitHub who ports doom to everything. What about that guy? Who's gonna verify him now? Google says that experienced users will still be able to install apps from wherever they want, even from unverified developers, although it'll take a few extra steps, a handful of warnings and swearing. You're not gonna blame Google when your phone mysteriously starts speaking in tongues. Nunteos technological spectas. I don't think anyone's gonna understand that at all. It's Latin. It's like a alternate universe where the holy spirit is malware. Yeah. According to Google's blog post, the idea is to curb malware without completely shutting the door on Android's openness, which, which has always been one of its defining features. The change is a response to heavy pushback from developers, industry groups, and everyday users who didn't love Google deciding for them what trusted means. So yeah, sideloading is back, baby. Just with training wheels. And when those come off, I guess that's when we really learn what malware feels like. Like a concrete skin burn. Skin burn. What do you call road rash? Road rash. That's how I learned not to longboard down hills with no protective equipment. Vine, the 6767 video app. Oh man, we have to leave that in vine, the 6 second video app shut down by Twitter in 2017, is back at it again at Krispy Kreme in the form of a new app called Divine, Complete with a sizable chunk of the original video library. The app was created by former Twitter employee Evan Henshaw Plath, AKA Rabble, while working for Twitter co founder Jack Dorsey's Not For Profit and Other Stuff. He didn't work for Jack Dorsey's Not For Profit and also for Other Stuff. The Not For Profits called and Other Stuff. Okay, Using data previously stored by Archive Team, which is again a separate entity, not the same as the archive.org team. Okay, rabble was able to restore 150 to 200,000 original vines from 60,000 creators who can contact him if they want to take control of their reconstructed accounts. Wow. And it seems like he's hoping that they will. Divine breaks from its short form video brethren like TikTok by banning AI generated content. It'll use tech from the human rights nonprofit the Guardian Project to confirm that a clip was truly captured on a smartphone. And if Divine spots AI slop, it'll yell is that a weed? And call the police, not on a microwave. So with Divine when you say look at all those chickens, be assured that while you might not actually be pointing at chickens, you certainly won't be pointing at AI chickens. And that's all we want. At the end of the day, Anthropic is in the spotlight after alleging that Chinese state linked hackers used their Claude chatbot to to try and hack into 30 of what anthropic calls critical organizations. And some of those intrusions actually succeeded, supposedly. According to Anthropic, attackers fed Claude detailed prompts, giving it tasks like finding vulnerabilities on selected targets, exploiting them, and even harvesting data. Claude was apparently able to oblige them, but it also hallucinated portions of its results, meaning that the world's first AI orchestrated cyber espion campaign included completely fictional cybersecurity threats. It's like staging a heist with that one overconfident guy who points at every wall and says there's a safe behind that and there isn't. Anthropic claimed the attack was up to 90% autonomous, a figure that a ton of other researchers immediately questioned, arguing that white hat hackers haven't been able to coax Claude into doing anything close to to what's described here, so it's unlikely the spy ring actually reached near full autonomy. Still, the fact that an LLM could meaningfully assist at all is probably raising the blood pressure of IT admins across the world, especially given how friendly these chatbots seem to be when you talk to them. But at least for now, this is less like Claude the Black Hat and more like Claude the Dunce Captain. Get em Riley. Oh man, that was a burn. That's almost as sick as our sponsor Micro Center. Hey Arizona, your sports teams aren't doing so hot. I'm sorry about that. But guess what? It's okay because on November 7, Micro center had the grand opening of their Phoenix location. This new store ranks up there with Dobbins Lookout or the Musical Instrument Museum as the city's top landmark probably will be soon. The new location is huge news because Micro center rules. They just won PC Mag's 2025 awards for best Big Box Store, best Brick and Mortar Tech Retailer, and most impressively, best Tech Retailer overall. You have gotta check these guys out. If you can't visit one of Micro Center's rad physical locations but still wanna be distracted from your terrible sports team, don't fret, they've got you covered online. November's sales are ever changing and only last about a week at a time, so check the Micro center website regularly for some amazing deals throughout the entire month of November. Now if the long bits we just did are the original trilogy, then the quick bits are the prequels. There's a lot more stuff jammed in there and they're harder to understand. Firefox is building what it calls an AI window, which is like an AI browser mode that is kept separate from your normal tabs, kind of like a private incognito window. In it, you'll be able to chat with an AI assistant and get help while you browse, but only if you want, because you can totally ignore this feature and not use it if you're prejudiced against clankers. Mozilla is asking their community to help shape this feature by signing up to test it and then give feedback. Seems like Mozilla's trying to keep up with all the trending AI browsers while still protecting their users privacy. Like me in the shallow end at a pool party. I'm present, I'm there, but I don't want to drown. I'm sorry, there's a lot of kids over there and they can all grab your face at any one time. Microsoft has fixed an issue where some Windows 10 users couldn't enroll for extended security updates to get an extra year of free updates. But wait, don't you have to be enrolled in that program to get more updates in the first place? Is Microsoft trapping users in liminal space where they're both getting updates and blocked from getting updates at the same time? No. Thankfully this isn't a catch 22 situation, and every Windows 10 user can get this new fix, which will allow them to use the ESU enrollment wizard and collapse the quantum uncertainty and we're back in our home dimension. Very good. Good job, Microsoft. So kind of you to allow people to do the thing you said they could do. Tesla is recalling over 10,000 of its Powerwall 2 home battery units in the US because a third party battery cell defect can lead to overheating, smoke, or even fires. The company says it has remotely discharged nearly all affected units to make them safe and will replace them free of charge with certified installers reaching out to customers. Oh hello, certified installer. Is it getting hot in here or is that just me? What? My house is Battery's on fire. ChatGPT is rolling out group chats in select countries, letting users invite contacts to join their ChatGPT session. So you can use ChatGPT to actually chat with another human now? As long as an AI is there to supervise. This is great news, especially if you want someone to mediate your dysfunctional family arguments about what to get for dinner tonight. Yeah, this is the what to do. Meanwhile, WhatsApp is launching third party chat integration across Europe, allowing fully encrypted conversations with people on other messaging platforms as part of the new requirements laid out by the Digital Markets Act. It doesn't look like ChatGPT is one of those now interoperable chat apps, so at least it won't butt in and offer to write some erotica or something. Which is always. It always does this. So annoying. Open ChatGPT. Good morning. Want some erotica? Stop. Stop asking. And a heal. A heal A new smart bandage uses an AI powered camera to take pictures of a wound every couple of hours, figure out how it's healing, and then either zap it with tiny pulses of electricity or release drugs to help speed the healing up. Which seems better, more useful to me. How's that healing? Oh, it's all right. What are you? Trust me, I'm a doctor. In tests on pigs, the bandage doubled the growth of new skin and cut inflammation. However, the effects are still modest and there's more work to do before it's ready for people. Animal testing can be a touchy subject, but that pig looks comfortable enough and cool in their little healing vest, so totally worth it. So any animal testing is actually fine as long as they look cool while they're doing cool. Yeah, that's actually how we found out cigarettes are fine because the animals look like badasses. Those are the coolest monkeys I've ever seen. And I bet whoever's hosting on Monday will look cool, too. How could they not? So make sure to come back. I'm gonna take the weekend to look into the Tech News sequels. I haven't seen them yet, but I bet they have to be meticulously plotted. What else would you do with a giant, important franchise? No way. They'll leave me disappointed in questioning my lifelong love for tech news. This'll be fine.