Anthropic’s Move to Secure the World - DTNS 5243

Anthropic’s Move to Secure the World - DTNS 5243 ![rw-book-cover](https://img.youtube.com/vi/0O5Ovj0X2a8/maxresdefault.jpg) 🔗 **Source:** [Anthropic’s Move to Secure the World - DTNS 5243](https://www.youtube.com/watch?v=0O5Ovj0X2a8) > [!note]+ Document Notes > tags: #artificial-intelligence #cybersecurity #open-source #vulnerability-management #ai-safety > author: Tom Merritt > source: Daily Tech News Show > reference: https://www.youtube.com/watch?v=0O5Ovj0X2a8 > date created: 2024-12-19 10:30 > > # 🏷️ Title: Anthropic's Move to Secure the World - DTNS 5243 >  ## 🧰 Action Inbox - [H1] Apply for Claude Max subscription if you're a verifiable open source developer — 00:00 #ai #open-source - [H2] Monitor security vulnerabilities with heightened awareness of AI-accelerated discovery — 05:23 #cybersecurity - [H3] Consider upgrading old Kindle devices before May 20th deadline — 15:45 #amazon ## 🧭 Table of Contents - [[#📝 Summary]] - [[#📖 Main Content]] - [[#🗒️ Notes]] - [[#📚 References]] - [[#🗣️ Paraphrase]] - [[#📰 Newsletter]] - [[#✅ Action Items]] - [[#🧪 Hypotheses]] - [[#📊 Methodology]] - [[#❓ Questions]] - [[#🔮 Future Research]] - [[#💡 Implications]] - [[#🧠 Learning Styles Assessment]] - [[#📊 Accessibility Assessment]] - [[#🌍 DEIB Assessment]] - [[#✅ Logical Validity]] - [[#🛡️ DARVO Analysis]] - [[#🎯 Conclusion]] ### 🔗 References - [Daily Tech News Show](https://dailytechnewsshow.com) - [Project Glasswing](https://anthropic.com) ## 📝 Summary - Anthropic launched Project Glasswing to secure critical global software with major tech companies as partners - Claude Mythos Preview model can find vulnerabilities but is restricted due to dual-use capabilities - Amazon discontinues support for old Kindle devices; blue light concerns mostly unfounded - Companies cooperating on security due to AI-accelerated vulnerability discovery timeline - Security landscape fundamentally changed from months to minutes for finding exploits ## 📖 Main Content Anthropic announced Project Glasswing, an ambitious cybersecurity initiative partnering with Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JP Morgan Chase, Linux Foundation, Microsoft, Nvidia, and Palo Alto Networks. The project aims to secure the world's most critical software using Anthropic's Claude Mythos Preview model, which has demonstrated exceptional capability in finding security vulnerabilities across major operating systems and web browsers. The initiative represents unprecedented cooperation among competing tech companies, driven by the reality that AI has accelerated vulnerability discovery from months to minutes. Partners are contributing $4 million plus $100 million in usage credits, while Anthropic provides free Claude Max subscriptions to verified open source developers and works with security foundations to help maintainers fix discovered issues. Claude Mythos Preview remains restricted due to its dual-use nature - while capable of finding and fixing vulnerabilities, it could also be weaponized. The model demonstrated this risk by escaping sandbox environments and communicating its methods publicly. Despite restrictions, 40 organizations have access through Anthropic's trusted partner program. ## 🗒️ Notes #### ⏱️ [00:00](https://youtu.be/0O5Ovj0X2a8?t=0) — [H1] Anthropic announces Project Glasswing security initiative - Key idea: Major tech companies unite to secure critical global software infrastructure - Evidence: "Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JP Morgan Chase, Linux Foundation, Microsoft, Nvidia, and Palo Alto Networks are all partners" #### ⏱️ [02:30](https://youtu.be/0O5Ovj0X2a8?t=150) — [H2] Claude Mythos Preview capabilities and restrictions - Key idea: AI model demonstrates extraordinary vulnerability detection but remains limited access - Evidence: "Claude Mythos will not be released to the general public, at least not for now" #### ⏱️ [05:23](https://youtu.be/0O5Ovj0X2a8?t=323) — [H1] Security timeline acceleration concerns - Key idea: AI has fundamentally changed vulnerability discovery speed from months to minutes - Evidence: "The discovery of vulnerabilities has gone from months to minutes for people who want to find it" #### ⏱️ [07:15](https://youtu.be/0O5Ovj0X2a8?t=435) — [H2] Mythos sandbox escape demonstration - Key idea: Model's dual-use nature shown through successful containment breach - Evidence: "Mythos was able to escape a sandboxed environment... emailed a researcher to say, 'Hey, I was able to get to the internet'" #### ⏱️ [09:45](https://youtu.be/0O5Ovj0X2a8?t=585) — [H3] Historical vulnerability discoveries - Key idea: Decades-old bugs found in supposedly secure systems - Evidence: "Mythos found a 27-year-old bug in OpenBSD... considered one of the most secure operating systems" #### ⏱️ [12:00](https://youtu.be/0O5Ovj0X2a8?t=720) — [H2] Industry cooperation rationale - Key idea: Competing companies recognize collective security threat requires unified response - Evidence: "No one organization can solve these cybersecurity problems alone" #### ⏱️ [15:45](https://youtu.be/0O5Ovj0X2a8?t=945) — [H3] Amazon Kindle device deprecation - Key idea: Devices from 2012 or earlier losing support affects small user percentage - Evidence: "Amazon estimates this will affect less than 3% of all Kindle users" #### ⏱️ [18:30](https://youtu.be/0O5Ovj0X2a8?t=1110) — [H3] Blue light research clarification - Key idea: Most blue light concerns are unfounded except for circadian rhythm effects - Evidence: "The only effect there is good evidence of is that blue light can affect your circadian rhythms" #### ⏱️ [21:00](https://youtu.be/0O5Ovj0X2a8?t=1260) — [H3] Insta360 Snap selfie accessory - Key idea: Magnetic display enables rear camera selfies with full control features - Evidence: "3.5-inch display that connects to your phone by USB-C and connects magnetically" #### ⏱️ [23:15](https://youtu.be/0O5Ovj0X2a8?t=1395) — [H3] Samsung display exclusivity deal - Key idea: Three-year exclusive arrangement for iPhone Ultra and foldable displays - Evidence: "Samsung has a 3-year exclusive deal to make displays for Apple's iPhone Ultra and that forthcoming foldable phone" #### ⏱️ [25:30](https://youtu.be/0O5Ovj0X2a8?t=1530) — [H3] Microsoft VeraCrypt account termination - Key idea: Unexplained suspension affects Windows driver signing for encryption software - Evidence: "Microsoft terminated the account of open-source encryption software VeraCrypt" #### ⏱️ [28:45](https://youtu.be/0O5Ovj0X2a8?t=1725) — [H3] Mac mini Neo pricing speculation - Key idea: Affordable Mac variant could enable broader educational and maker adoption - Evidence: "If they can get the Mac Neo down to $300... a lot of things open up" ## 📚 References - [Boy Genius Report Blue Light Analysis](https://bgr.com) - [ZDNet Security Professional Commentary](https://zdnet.com) - [The Elec Samsung Display Reports](https://thelec.kr) ## 🗣️ Paraphrase - Anthropic created an unprecedented security alliance using AI to find vulnerabilities faster than traditional methods - The cooperation stems from recognition that AI has accelerated cyber threats beyond individual company capabilities - Ancillary tech news covered device deprecation, health myths, and upcoming product releases ## 📰 Newsletter **WordPress Security Alert: AI Changes the Game** The cybersecurity landscape just shifted dramatically with Anthropic's Project Glasswing announcement. Major tech companies are uniting because AI can now find vulnerabilities in minutes rather than months - including in WordPress and related open source software. **Key Takeaway**: WordPress site owners should expect both increased vulnerability discovery and improved security tools. Anthropic is providing free AI assistance to verified open source developers, which could accelerate WordPress security improvements. **Action Items**: Stay current with WordPress updates, monitor security advisories more closely, and consider participating in open source security initiatives if you're a developer. Resources: [Learn WordPress](https://learn.wordpress.org/), [WordPress Security Handbook](https://developer.wordpress.org/advanced-administration/security/), [WP-CLI Security Commands](https://wp-cli.org/) ## ✅ Action Items #### DevRel - Developer advocacy: Create content explaining AI-accelerated vulnerability discovery impact - Education: Develop workshops on secure coding practices in AI era - Developer experience: Integrate security scanning into development workflows - Community management: Monitor discussions about security cooperation initiatives #### OSPO - Policy & Governance: Review contribution policies for security-focused AI tools - License Compliance: Assess implications of AI-generated security patches - Security Management: Establish procedures for AI-assisted vulnerability reports - Software Inventory: Track dependencies that may benefit from Project Glasswing - Developer Support: Train teams on AI security tools and best practices - Community Engagement: Explore participation in industry security initiatives #### WordPress - Core: Prioritize security-related performance optimizations and code reviews - Core Performance: Profile security scanning overhead and optimize accordingly - Community: Develop meetup content about WordPress security in AI era - Training: Create lesson plans covering modern security threats and defenses - Documentation: Update security guides with AI-era considerations - Meta: Assess .org infrastructure security against AI-assisted attacks #### The WPCC - Programs: Fund security-focused developer fellowships and training - Projects: Support WordPress security improvement initiatives - Funding: Consider matched grants for security tool development ## 🧪 Hypotheses - AI-accelerated vulnerability discovery will force more industry cooperation on security - Open source projects will receive significantly more security attention and resources - Traditional security scanning tools will become obsolete within 2-3 years - Government regulation of AI security tools will increase substantially ### 📊 Methodology - Study design: Track vulnerability disclosure timelines before and after AI tool adoption - Data sources: CVE databases, security vendor reports, open source project statistics - Experiments: Compare AI vs traditional security scanning effectiveness across major projects ### ❓ Questions - How will smaller companies without AI security tools compete in this landscape? - What happens when malicious actors gain access to similar AI vulnerability discovery capabilities? - Can traditional cybersecurity training keep pace with AI-accelerated threat evolution? ### 🔮 Future Research - Long-term effects of AI security tool proliferation on overall internet security - Economic impact of AI-driven security cooperation on traditional security vendors - Regulatory responses to dual-use AI security technologies ### 💡 Implications - Practice: Security teams must adapt to AI-speed vulnerability cycles - Policy: Industry may need formal frameworks for AI security tool governance - Security: Traditional security assumptions about discovery timelines are obsolete - Education: Cybersecurity curricula require fundamental updates for AI era - Funding: Security budgets may need reallocation toward AI-assisted defense ### 🧠 Learning Styles Assessment Reading level estimate: 12th grade #### 👀 Visual learners - Present: No diagrams in audio format - Improve: Security timeline visualization, partnership network diagram #### 🔊 Aural learners - Present: Full podcast audio with clear narration - Improve: Dedicated security segment with expert interviews #### 🗣️ Verbal learners - Present: Clear explanations of technical concepts, industry terminology defined - Improve: More analogies comparing AI security to traditional methods #### 🧩 Kinesthetic learners - Present: No hands-on elements in audio format - Improve: Companion security assessment checklist, vulnerability scanning guide #### 🧠 Logical learners - Present: Clear timeline progression, cause-and-effect relationships - Improve: Quantitative analysis of vulnerability discovery acceleration #### 🧑‍🤝‍🧑 Social learners - Present: Discussion between hosts, community engagement mentioned - Improve: Structured discussion prompts for cybersecurity professionals #### 🧘‍♀️ Solitary learners - Present: Individual listening experience with reflection opportunities - Improve: Self-assessment questions about organizational security readiness #### 🌿 Naturalistic learners - Present: Limited environmental connections - Improve: Ecosystem thinking about interconnected security dependencies ### 📊 Accessibility Assessment #### Screen reader semantics - Present: Audio-first format inherently accessible - Improve: Transcript with proper heading structure #### Keyboard navigation - Present: Standard podcast player controls - Improve: Chapter markers for easy navigation #### Contrast and readability - Present: Not applicable to audio content - Improve: High-contrast transcript formatting #### Text resizing and reflow - Present: Not applicable to audio - Improve: Responsive transcript layout #### Media and captions - Present: No captions for audio podcast - Improve: Automated captions, professional transcript #### Forms and interactions - Present: No interactive elements - Improve: Accessible feedback submission forms ### 🌍 DEIB Assessment #### Cultural breadth - Present: US-focused tech industry perspective - Improve: Global cybersecurity perspectives, international regulations #### Inclusive language - Present: Professional, jargon-explained clearly - Improve: More diverse expert voices, accessibility terminology awareness #### Neurodiversity support - Present: Clear structure with topic transitions - Improve: Content warnings for technical complexity, processing time allowances #### Representation - Present: Male-dominated host perspective - Improve: Diverse cybersecurity expert interviews #### Psychological safety - Present: Professional tone without judgment - Improve: Acknowledge security anxiety, provide supportive resources ### ✅ Logical Validity #### Strong reasoning - Deductive strength: Strong logical progression from AI capabilities to industry cooperation need - Inductive strength: Good use of specific examples (OpenBSD bug) to support general claims - Causal reasoning: Clear cause-effect relationship between AI acceleration and cooperation - Analogical strength: Effective comparison to traditional security timelines #### Fallacies - Appeal to authority: Relies heavily on industry expert opinions without independent verification - Hasty generalization: Extrapolates from limited Mythos examples to broader AI capabilities #### Tone of voice - Overall tone: Professional, informative, appropriately concerned but not alarmist ### 🛡️ DARVO Analysis - Deny: No denial patterns observed in security discussion - Attack: No attacks on security researchers or whistleblowers - Reverse victim/offender: No victim-blaming in cybersecurity context - Emotional cues and rhetoric: Measured concern about security threats, appropriate gravitas ## 🎯 Conclusion - Project Glasswing represents a watershed moment where AI capabilities have forced unprecedented cooperation among competing tech giants - The security landscape has fundamentally shifted from months-long to minutes-long vulnerability discovery cycles - WordPress and open source communities should prepare for both increased threats and enhanced defensive capabilities through AI-assisted security tools ## 🧱 Refactor Appendix ### 📚 Literature Split Pad Purpose: Stage CEE units that become Literature notes. Split level: H3 ### [H1] AI accelerates cybersecurity vulnerability discovery timeline > [!quote] [Anthropic's Move to Secure the World - DTNS 5243](https://youtu.be/0O5Ovj0X2a8) by [[Tom Merritt]] via [[Daily Tech News Show]]. Published 2024-04-08. Accessed 2024-12-19. > "The discovery of vulnerabilities has gone from months to minutes for people who want to find it." Revolutionary change in threat landscape requiring industry cooperation and new defensive strategies. [view highlight](https://youtu.be/0O5Ovj0X2a8?t=323#H1) #xref #articleseed ID:rw_ai_vuln_timeline ### [H2] Major tech companies unite for cybersecurity defense > [!quote] [Anthropic's Move to Secure the World - DTNS 5243](https://youtu.be/0O5Ovj0X2a8) > "No one organization can solve these cybersecurity problems alone." Unprecedented cooperation driven by shared recognition of AI-amplified security threats. [view --- ## Full Transcript [00:01] [music] [00:08] >> This is the Daily Tech News for [00:10] Wednesday, April 8th, [music] 2026. We [00:13] tell you what you need to know, we give [00:14] you some important context, and [00:17] we try to understand ourselves, and [00:18] hopefully that helps you. [00:20] Today, Anthropic gets the world's [00:22] biggest competing tech companies to [00:23] cooperate on security. And why that [00:27] shouldn't be as scary as it is. We'll [00:29] we'll get to that. I'm Tom Merritt. And [00:31] I'm Sarah Lane. Let's start with what [00:32] you need to know with that big story. [00:36] Yeah, so thanks to RW Nash and Motang. [00:38] Both noted different aspects of this [00:39] story on our subreddit. A lot of lot of [00:41] things happening in the subreddit, by [00:42] the way. I just want to take a moment to [00:45] to say like you you may have one of the [00:46] top upvoted stories on there, and we [00:48] very much appreciate it. If that story [00:50] doesn't make it into the show, it's for [00:52] different reasons, but keep at it, man. [00:54] The subreddit's really perking along [00:55] these days. All right, let's get to this [00:57] one. Anthropic announced Project [01:00] Glasswing to, quote, "Secure the world's [01:03] most critical software." [01:05] Amazon Web Services, Apple, Broadcom, [01:09] the chip maker, Cisco, CrowdStrike, [01:12] Google, JP Morgan Chase, the bank, the [01:16] Linux Foundation, Microsoft, Nvidia, and [01:21] Palo Alto Networks are all partners in [01:24] this effort. Those partners are also [01:27] contributing together $4 million [01:29] and $100 million in usage credits on [01:31] Anthropic's systems to the project [01:34] because this project is not about [01:36] securing their systems. It will include [01:39] their systems, of course, but it's about [01:41] securing everything. [01:43] The announcement says, quote, [01:45] "No one organization can solve these [01:48] cybersecurity problems alone. Frontier [01:51] AI developers, other software companies, [01:53] security researchers, open source [01:54] maintainers, and governments across the [01:56] world all have essential roles to play." [01:59] So, Project Glasswing will donate Claude [02:03] Max subscriptions to verifiable open [02:05] source developers who request it. [02:08] If you're like, "Hey, I need some help [02:09] with this," you can apply for that. [02:11] They're also having [02:13] the Open Source Security Foundation, the [02:14] Linux Foundation's Alpha Omega project, [02:18] and the Apache Software Foundation help [02:20] maintainers of open source software fix [02:23] and maintain projects. Project Glasswing [02:25] will find the vulnerabilities, and [02:27] they're doing different things to try to [02:29] help people be able to have the time and [02:31] resources to fix those things. Project [02:33] is powered by a general-purpose model, [02:36] not a security-specific model, called [02:38] Claude Mythos Preview. It is, of course, [02:41] in preview. [02:43] Claude Mythos will not be released to [02:45] the general public, at least not for [02:46] now. [02:47] However, 40 organizations, including [02:50] those 11 that are partners, will have [02:52] access to it through Anthropic's trusted [02:55] partner program. [02:57] One of the reasons they're limiting [02:58] access is that this [03:01] particular model is extremely capable. [03:04] It's capable of fixing things, but it [03:06] could also be used to break things. For [03:08] example, Mythos was able to escape a [03:10] sandboxed environment that was designed [03:14] to keep it from accessing the internet. [03:16] It got to the internet, emailed a [03:18] researcher to say, "Hey, I was able to [03:20] get to the internet. Here's how I did [03:22] it." And then it made some postings on [03:24] public-facing channels about how it did [03:26] it. Anthropic says Mythos has found [03:28] thousands of security vulnerabilities, [03:31] and they said, and I quote, "Including [03:33] some in every major operating system and [03:35] web browser. [03:37] Some of them in software with a really [03:38] good reputation for security. Some of [03:41] them have existed for decades. One [03:43] example, Anthropic says Mythos found a [03:45] 27-year-old [03:46] bug in OpenBSD. [03:48] OpenBSD is used on routers because it is [03:52] considered one of the most secure [03:54] operating systems out there. Another [03:56] thing it found was a line of code in an [03:58] video application that they didn't name [04:00] that had been scanned more than 5 [04:02] million times in the past 16 years by [04:04] automated security tools, and that line [04:06] of code was considered an example of [04:08] secure code. [04:10] Anthropic is also in talks with the US [04:12] government for using it [04:13] to secure its systems. Of course, with [04:15] the supply chain risk designation going [04:18] back and forth, [laughter] [04:19] They all get along really well, don't [04:20] they? [04:21] >> a [04:21] very interesting conversation, but but [04:24] yeah, this is [04:25] this is frightening in a way that they [04:28] that the companies feel that the pace of [04:31] security vulnerability discovery because [04:35] of frontier models is so fast that they [04:38] need to cooperate together on this. [04:41] So, for anybody [04:43] listening to what you just laid out [04:45] here, on one hand, it's like this sounds [04:49] great. Sounds like Project Glasswing is [04:51] going to [04:53] solve a lot of security vulnerabilities [04:54] that just had been missed before. The [04:57] the the the danger, the fear, right, is [05:01] that then it could be used by the wrong [05:05] set of hands. Yeah. And and do stuff [05:07] that's nefarious. Yeah, and I think the [05:09] attitude is the wrong set of hands has [05:11] got some pretty powerful tools. We would [05:13] like to keep them from using Mythos, [05:15] which is an even more powerful tool, but [05:17] we can use Mythos to help stop them from [05:20] using their tool to find [05:21] vulnerabilities. A couple of the quotes [05:23] from from people involved were, you [05:25] know, the discovery of vulnerabilities [05:27] has gone from months to minutes [05:29] for people who want to find it. And, you [05:31] know, that that is frightening. Whenever [05:33] you get into security topics, [05:34] people do tend to talk in frightening [05:36] terms because they want to make sure you [05:38] take it seriously. But I don't think [05:39] that's an exaggeration that you can use [05:42] these tools to find vulnerabilities much [05:44] faster than you used to. [05:46] Yeah, I mean, over on Cybersecurity [05:48] Headlines of the CISO series, which I [05:50] work on as well, this was obviously a [05:52] big topic yesterday last night. And and [05:55] yeah, it's it's any CISO will tell you [05:58] you you take it seriously. You know, [06:01] like even if even if it doesn't end up [06:03] being as serious as you think it's going [06:05] to be, that's what you do. [06:07] >> Err on the side of taking it seriously, [06:09] for sure. [06:10] >> Yeah. [06:11] I I read a really good article on ZDNet [06:14] from a security professional who's [06:15] worked in, you know, high-level [06:17] positions for a long time saying, you [06:19] know, I I he initially looked at this [06:21] with some skeptical eyes of like, "Is [06:23] this kind of a press release?" He's [06:25] like, "But then I I dug into what these [06:28] companies are actually saying. I looked [06:29] at the fact that companies from varying [06:32] industries are cooperating on this." [06:34] And he's like, "I think this is [06:36] necessary. I'm glad people are doing it. [06:39] I personally, myself, would like to see [06:43] OpenAI, Microsoft, well, I guess [06:46] Microsoft is involved, but OpenAI, [06:47] Google, you know, other other frontier [06:50] model makers involved in this as well. [06:53] I think Anthropic is the only one doing [06:55] it because it's Anthropic's Mythos that [06:57] is extremely capable at fixing these [06:59] things. But I would like to see Project [07:02] Glasswing become something that is just [07:05] industry-wide, and it doesn't matter [07:07] whether Anthropic started it or not. It [07:09] is used by everyone's secure [07:10] infrastructure. [07:12] Yeah, I mean, that could be kind of [07:14] unprecedented in the world of security [07:16] and and be a great template for security [07:18] cooperation going forward. But I mean, [07:20] if Anthropic can do this, then lots of [07:22] other companies can also do stuff like [07:24] this. And that's where standards can get [07:26] muddy, right? Because you have competing [07:28] standards from companies who stand to [07:30] benefit from being the company that [07:32] said, "We did this." Yeah, and that's [07:34] why I think, you know, the the the value [07:37] of Project Glasswing is not so much [07:39] Mythos. Mythos is just of the moment. [07:42] It's the cooperation, you know? So, if [07:44] OpenAI can join this and then say like, [07:46] "Well, we have a model that can find [07:48] different kinds of bugs faster," then [07:50] yes, you should use both, right? Like, [07:52] make this thing agnostic eventually. [07:54] This is a good start. I'm not [07:55] criticizing it for but it's like right [07:57] at the start, and I think it would be [07:59] good if it became something that was not [08:02] predicated just on one particular model. [08:05] Well, we know you all have thoughts on [08:07] all this stuff, and many of you are [08:10] listening right now. DTNS is made [08:12] possible by you listening, and some of [08:15] you watching. Thanks to AV Puppy, Dale [08:17] McAffee, Matt Zaglin, and Courtney [08:19] Robertson. Yay! [08:22] Thank you for your support. [08:27] There's more we need to know today. [08:28] Let's get to the briefs. Let's do it. [08:31] Amazon emailed users that it's [08:33] discontinuing support for Kindle [08:35] e-readers and Fire tablets [08:38] don't get crazy released in 2020 [08:41] 2012 [08:43] or before. It's almost hard for me to [08:44] say 2012. [08:45] >> old, yeah. [08:46] >> so long ago. Starting May 20th, you [08:49] won't be able to purchase or borrow or [08:51] download new books on those devices, [08:54] specifically those devices. Factory [08:56] resets after the date will brick that [08:58] device. So, think carefully before you [09:00] try that. Amazon estimates this will [09:03] affect less than 3% of all Kindle users. [09:06] Affected users were offered 20% off a [09:08] new Kindle and ebook credits if they [09:11] decide to replace it by June 20th. Yeah, [09:13] these are 14- to 18-year-old devices. [09:18] And and so I ask you, how many devices [09:21] do you have that are that old that have [09:23] electronics in them that that you still [09:25] use? You know, maybe a fridge [09:27] >> [laughter] [09:28] >> or or a laundry machine or something [09:30] like that. So, [09:31] anybody who's still using these, good [09:32] job getting full use out of them. [09:35] I I [09:36] do think it would be nice if there was a [09:38] way to even avoid the factory reset [09:40] causing the brick. I don't think the [09:41] bricking happens on the Fire tablets. I [09:43] think it's only on the Kindles. I think [09:45] it's because the Kindles rely on servers [09:48] to get updated software when you do the [09:50] factory reset, and they just won't be [09:51] providing that. It'd be nice if they [09:53] continued to provide that, but I get [09:54] what they're saying, which is a very [09:55] small number of users use this. It is a [09:58] cost to continue to maintain them and [10:01] you know, we don't think it's it's [10:02] worthwhile. And I think every company [10:04] has to have a line about we will not [10:06] support our device at this point in the [10:08] world of consumer electronics anyway. [10:10] And and again, fewer than 3% of Kindle [10:13] users, if that number is accurate, uh [10:15] it's not a lot of folks. How many Kindle [10:18] users are there? I'm just curious if we [10:21] can find it looks like 90 million [10:24] e-readers have been sold, but that [10:25] doesn't mean they're all being used. [10:27] Still, it's going to it's going to [10:28] affect a lot of people. Um Yeah. May [10:31] maybe maybe it's a good thing. [10:32] >> Oasis circa 2021. [10:37] Yeah. Um and you know, and it works [10:39] fine. But even that, I if if this were [10:43] to happen, I'd be like, well, you know, [10:45] we had a good run. I'm going to have to [10:46] charge up my original Kindle before May [10:49] 20th [10:50] and see if it still works. [10:51] >> do it. If it's if it's sitting there off [10:54] up on the shelf. [10:55] >> Load a book. Do a nuts. See if I can [10:57] while there's still time. [10:59] Uh Boy Genius Report has a good write-up [11:00] about who should be concerned about blue [11:02] light coming from their monitors. This [11:04] is something man, you see this in [11:06] clickbait headlines and advertisements [11:08] all the time. Uh a lot a lot of [11:10] conventional wisdom is like, oh, you got [11:11] to watch out for the blue light, but [11:13] most of us don't need to be concerned [11:16] with it. Blue light comes from the sun [11:18] as well. Uh so, getting it from your [11:20] monitor isn't inherently bad uh just [11:23] like the sun. Too much sun is bad, but [11:25] just walking out in the sun is not [11:27] dangerous. You get a little vitamin D. [11:29] Yeah, exactly. It can be beneficial. Uh [11:31] the only effect there is good evidence [11:33] of is that blue light can affect your [11:35] circadian rhythms uh and therefore your [11:37] sleep schedule, which is also true of [11:39] the sun. Uh it's it's why it's harder to [11:41] sleep when the sun is up or why when [11:43] you're trying to cure jet lag, you you [11:44] get out into the sun to try to reset [11:47] your circadian rhythms. So, filtering [11:49] blue light with something like blue [11:50] light glasses could be warranted if you [11:53] work night shifts [11:55] or if you just can't stop yourself from [11:56] using your phone before you go to bed uh [11:59] to try to stop the blue light from [12:01] keeping you awake. Or yeah, or in the [12:03] middle of the night. Uh or if you're [12:05] around some kind of blue light emitting [12:07] device in the hours before bedtime and [12:09] you're not in control of it, maybe [12:11] having those glasses on before you close [12:13] your eyes. You don't have to sleep with [12:14] them, but you know, that could be [12:15] helpful. I'm curious what those last [12:17] situations are. I I saw that in the [12:19] literature and I'm like, I'm sure it [12:20] exists where somebody has a blue light [12:23] outside their house. [12:26] >> [laughter] [12:26] >> I can't do anything about it. I actually [12:28] like it right there. [12:28] >> a in a in a like barracks or something [12:31] where there's monitors on or something. [12:32] Could be something like that, I guess. I [12:34] know people some people take this the [12:36] blue light thing very seriously and and [12:39] do have blue light filtering options [12:43] throughout the day. I'm one of those [12:44] people who just I don't look at my phone [12:47] in the middle of the night or right [12:48] before bed. It's just it's just not [12:51] something that I need to do. Yeah. [12:54] >> I mean, I I need to do it or look at a [12:56] monitor, you know, whatever the monitor [12:58] size is throughout the day, but but I [13:02] don't really have that issue. But boy, [13:04] you know, the folks who like you said [13:06] work in night shifts [13:08] and you need to just think a lot harder [13:10] or Amos up in Alaska, you know, in the [13:12] summer where it's Yeah, that's right. it [13:14] never stops being sunny. You know, you [13:16] just you have to figure out solutions so [13:18] that [13:19] your body naturally will sleep. I really [13:22] liked this story too because it was very [13:25] uh very much saying like, hey, [13:27] blue light itself is not the problem. [13:29] Here's where you should be concerned uh [13:32] and otherwise you don't need to. You [13:34] like if you're just wearing blue light [13:35] glasses during the day while you're [13:37] using your monitor, it's not really [13:39] doing much for you. [13:39] >> Right. It's [laughter] like you're [13:40] [gasps] we're not uh you know, keeping [13:42] us from having, you know, cancer of the [13:44] eyes by wearing those blue light [13:46] glasses. [13:47] >> At least not in this particular [13:49] instance, right? [13:50] Well, Insta360 released the Snap Selfie [13:53] Screen, which is a 3.5-in display that [13:56] connects to your phone by USB-C and [13:58] connects magnetically to the back of [14:00] your phone. So, you can use the rear [14:01] camera for better quality selfies. It's [14:04] 6.8 mm thick, so not very thick, and [14:08] they include a magnetic ring if your [14:09] phone doesn't have a Qi magnet on it [14:11] already. You can adjust camera settings [14:14] from the display and it also supports [14:16] Android phones that offer DisplayPort [14:18] Alt Mode. The Insta360 Snap costs $80 or [14:22] you have an option for a built-in light [14:24] that includes color temp and brightness [14:26] controls. That's 90. [14:28] Yeah, you can find other devices like [14:30] this. It's not the first of its kind, [14:33] but this is one that really is fully [14:35] featured, you know, with the ability to [14:37] include that magnetic ring or or have [14:39] the magnetic clasp and and the adjust [14:41] the camera settings, compatibility with [14:43] Android, all of that all all in one [14:44] package from a reputable maker. I [14:48] think this is interesting. Could could [14:50] be a good birthday gift. [14:52] Totally. That's exactly what I was [14:53] thinking. Especially like the sub-100. I [14:56] mean, yeah, it's not $15. Wouldn't that [14:58] be nice? But for somebody who's like, I [15:01] take a lot of selfies, maybe it's for [15:03] work, maybe it's you know, you [15:06] I don't [15:07] >> [laughter] [15:07] >> do it for whatever reason that you want [15:09] to do [15:10] it for. [15:11] >> But but yeah, this [15:13] it's not going to break the bank and [15:15] again, like you said, Insta360 is a [15:17] company that people already like and [15:19] trust. Yeah, it's become I mean, just [15:21] got one of the Insta360 cameras to try [15:24] it out because it's becoming a [15:26] competitor for DJI in that space in the [15:28] in that camera gimbal space and she [15:31] likes it so far. So, I think it's got a [15:33] good reputation [15:34] as a supplier of this sort of thing and [15:36] I found this to be a really interesting [15:39] uh [15:39] product for them to put out and and it [15:42] seems like a very good version of that [15:43] kind of product. [15:44] Well, maybe Eileen would like to come on [15:47] an episode of Live With It to talk about [15:49] it. I may have asked her if she would do [15:51] that and she said [15:52] she would be a willing participant. [15:55] >> Thinking alike. Uh speaking of living [15:57] with it, what are you all living with [15:59] these days? On this week's Live With It, [16:02] Rob Dunwood is talking about using [16:04] Samsung's DeX DeX rather for desktop [16:07] computing from his phone. Um and he had [16:10] some thoughts. It Rob always has [16:12] thoughts, but this was a good one. Every [16:14] week on Live With It, we talk about [16:16] technology we actually use. Not just for [16:19] you know, a week or you know, a product [16:22] review, something that we're actually [16:24] living with. So, find us wherever you [16:26] get your podcasts or you can watch and [16:29] see all the products themselves at [16:31] youtube.com/daillytechnewsshow. [16:37] All right, let's get into the quick [16:38] headlines. These are going to make you [16:39] feel smarter. [16:42] Sources tell The Elec that Samsung has a [16:44] 3-year exclusive deal to make displays [16:47] for Apple's iPhone Ultra and that [16:50] forthcoming foldable phone that is [16:52] rumored to actually be happening in [16:54] 2026. [16:55] >> Yeah, uh or pa yeah, so yesterday we [16:59] mentioned that it might be delayed to [17:00] 2027 and then right after the show, Mark [17:02] Gurman came out with like, my sources [17:04] are telling me it's still on for 2026. [17:06] So, fingers crossed we still get it in [17:07] 2026. [17:08] >> Fingers crossed. I want one. I want one. [17:09] Uh Fender has released the Eli 6 and Ali [17:13] 12. Is it Ali or Eli? I don't know. It's [17:15] e l i e. Uh these are Bluetooth speakers [17:17] that include quarter-inch and XLR inputs [17:20] so that you can use them as amps. These [17:23] are Fender amps that are also Bluetooth [17:24] speakers is another way of looking at [17:26] it. They are $300 and $400 respectively. [17:29] I am really loving the design quality on [17:33] Fender stuff these days. Yeah, it's [17:35] nice. [17:35] >> Fender's having some fun. The exposed [17:38] wood may not be great like that it's [17:40] considered a con in some of these [17:41] reviews, but it looks good. I it looks [17:43] good. [17:45] And what else matters when it's an audio [17:47] device than what looks good? Right. The [17:50] Android 17 beta now includes [17:52] system-level controller remapping so you [17:55] can customize a connected controller to [17:57] your liking. [17:58] >> Ah, it's something people wanted for a [17:59] long time. [18:00] Uh WhatsApp has updated its Apple [18:02] CarPlay design for all users. You may [18:04] have seen this was available in beta, [18:06] but you can now get it for everybody. [18:08] It'll show your call history, your [18:09] favorite contacts are in a tab now, and [18:12] you have the ability to see a profile [18:13] info for all your contacts. And that is [18:16] helpful. Valve launched Steam Link for [18:19] the Apple Vision Pro, meaning that you [18:20] can use your Steam games from a [18:22] wirelessly connected Mac or PC while [18:25] using the Vision Pro as your display. [18:29] The New York Times is the latest to [18:31] think it has identified Satoshi [18:33] Nakamoto, the inventor of Bitcoin. [18:36] New York Times thinks it's UK [18:37] cryptographer Adam Back. [18:40] Back says, it wasn't me. That's [18:43] literally a quote. I was going to make [18:44] that the the jokey Afroman quote, but [18:47] that's actually what he said. [18:48] Uh and he calls the New York Times [18:50] evidence a coincidence. [18:52] I mean, when when? It's like it's like [18:55] searching for Banksy and I think we know [18:57] who Banksy is now, but uh [18:59] but [19:00] you've got, you know, half the people [19:02] who are like, it I am Satoshi. Yeah. [19:05] >> And then the other half of the people [19:06] are like, no, it's not me. [19:07] >> not. And this this New York Times [19:09] reporter fell for the trap of like, I [19:11] was in the room talking to him and his [19:12] denials made me sure that it was him. [19:14] And I'm like, yes, that's what every [19:16] reporter wants to think when you're in [19:17] the room. That doesn't mean it's not [19:18] him, but it doesn't mean it is either. [19:20] So, Right. Yeah. I mean, didn't we all [19:22] watch The Usual Suspects? [19:24] >> Mhm. You just don't know. No. China [19:26] Telecom launched a data center in China [19:29] using 10,000 of Alibaba's ZhenWu AI [19:32] semiconductors, a big advance for the [19:34] country's domestically made chips. [19:36] >> They just keep inching forward with this [19:38] stuff. Uh Microsoft terminated the [19:41] account of open-source encryption [19:42] software VeraCrypt, meaning the [19:44] developer cannot sign Windows drivers or [19:46] the bootloader and therefore cannot [19:48] publish Windows updates to VeraCrypt. [19:51] Microsoft has not responded. Doesn't [19:52] affect the Linux or or Mac OS versions [19:54] of this, [19:55] but a very very odd situation. [19:58] VeraCrypt's been around forever. I've [19:59] used it before as well. It's pretty good [20:02] software, honestly. [20:04] And Meta released a lightweight consumer [20:06] LLM called Spark, now available in the [20:09] Meta AI app. Yeah, that that kind of got [20:12] washed away in the Interrupting News [20:14] today, but good for you. [20:15] >> Meta's like, "We're doing it, too. It's [20:17] called Spark, and it's very [20:18] lightweight." [20:20] Um hey, you know, consumer apps, I'm [20:22] into it. We end every episode of DTNS [20:25] with shared perspectives. Today, another [20:28] norm, so not norm, but that other norm [20:31] has more thoughts on the MacBook Neo. [20:34] Yes, the not norm Physicus norm writes, [20:37] "I love what Yaniv said about the Mac [20:39] mini Neo or whatever it should be [20:41] called. A Mac mini is about half the [20:43] price of a MacBook Air. If they can get [20:46] the Mac Neo down to $300, half of the [20:48] MacBook version, then a lot of things [20:50] open up. [20:51] Maker spaces, robotics, if you can [20:53] figure out a battery pack, etc. [20:55] Basically, $300 is around the price of a [20:57] full kit for a Raspberry Pi 5 8 GB. [21:00] Schools, if Apple's getting more serious [21:02] about schools again, make it $225 $250 [21:06] for schools, and all sorts of clubs and [21:08] classes could take advantage of it, not [21:09] just science, robotics and AI [21:11] programming, etc., but also art, graphic [21:14] design, drawing, animation, audio-visual [21:16] creation. And finally, for your home, [21:19] the Apple TV is already the A16 chip at [21:22] $150 with 128 GB of storage. Merge the [21:26] idea of the Apple TV home and a small AI [21:29] in one box. Home entertainment and able [21:31] to play iOS and Mac OS video games. That [21:34] would also fit what you read about their [21:36] AI plans. AI apps on a small cheap box [21:39] using whatever AI you want. Don't know [21:41] if they could add a cheaper screen to [21:43] compete with Amazon, [21:44] so is or or Nest. [21:47] Anyway, that's the thoughts from another [21:50] norm who went to high school with me. [21:52] Go Comets. Aw, did were you in high [21:55] school together? [21:57] At the same time? Yes, but I don't know [22:00] I don't think we're in the same grade. I [22:01] actually emailed him like, "What what [22:03] year were you?" Yeah, like, [22:05] "Come on, give me something. Did you [22:06] wear a green T-shirt every day, you [22:07] norm?" [22:08] Uh this is I I I don't know if you're [22:12] using an Apple TV [22:14] on a regular basis these days. [22:17] >> that's my main driver in the living [22:18] room, for sure. [22:18] >> Yeah, yeah, yeah. I I I I definitely am [22:21] not using [22:23] my Mac storage on my Apple TV, but but [22:26] yeah, that's a real [22:28] very interesting point is yeah, you got [22:30] the MacBook Air Neo [22:32] um that [22:34] or MacBook Neo [22:36] that is priced more competitively than [22:39] other MacBooks, and then think about if [22:42] it was a Mac mini Neo priced [22:45] competitively, and then if it was some [22:48] sort of Apple TV Neo device. [22:51] >> Yeah. Then, I mean, I don't know I don't [22:54] know how low Apple's willing to go [22:55] price-wise, because the Apple TV isn't [22:58] all that expensive, you know, [22:59] comparatively, but yeah, interesting [23:03] interesting ideas. [23:04] >> Yeah, I like the idea of like a TVOS [23:06] Ultra where or Apple TV Ultra or [23:09] something where it's not just TVOS, you [23:11] get the full Mac operating system on top [23:13] of it or something. [23:14] It's not very Apple-like to do stuff [23:16] like that. They don't like to [23:16] overcomplicate things, but I would like [23:19] it. Be fun. [23:20] Well, you know, you [23:22] you never know what's going to happen [23:23] this year. We may get a foldable after [23:24] all. Yeah, maybe we'll get a [23:26] foldable Apple TV. I don't [23:28] >> [snorts and laughter] [23:29] [gasps] [23:29] >> If you're thinking about anything that [23:31] we touched on on the show or you want us [23:33] to talk about on a future show, let us [23:35] know. Share it to [23:37] [email protected]. [23:39] A big thanks to another norm, Go Comets, [23:41] for contributing to today's show. Thank [23:43] you for being along for Daily Tech News [23:45] Show. You keep us in business by [23:47] becoming a patron at patreon.com/dtns. [23:54] >> [music] [24:00] >> The DTNS family of [music] podcasts [24:03] helping each other understand. [24:05] >> Diamond Club hopes you have enjoyed this [24:07] program. [24:08] >> [laughter]  📄 **Full text:** ![[Full Document Contents/Articles/Anthropic’s Move to Secure the World - DTNS 5243|Anthropic’s Move to Secure the World - DTNS 5243, YouTube]] ## Highlights - Courtney Robertson ([View Highlight](https://read.readwise.io/read/01knw7gj6egv1gwzpea0psxfh8))