Top HN Weekly Digest · W16, Apr 13-19, 2026

A weekly Hacker News digest for readers who want the strongest stories and discussions from the entire week in one place.


0. Claude Opus 4.7 (anthropic.com)

1952 points · 1443 comments · by meetpateltech

Anthropic has released Claude Opus 4.7, featuring significant improvements in software engineering, instruction following, and high-resolution vision. The model introduces new "xhigh" effort controls and advanced cybersecurity safeguards while maintaining the same pricing as its predecessor, Opus 4.6. [src]

The release of Claude Opus 4.7 has sparked confusion and frustration among users regarding the new "adaptive thinking" feature, which some find difficult to configure and others blame for a perceived decline in model performance [0][7][8]. While the model demonstrates improved self-awareness regarding its own logical fallacies—such as failing to realize a car must be driven to a car wash—users report significant issues with hallucinations, overly restrictive cybersecurity filters, and a lack of transparency from Anthropic regarding capacity constraints [1][5][9]. Consequently, some developers are migrating to competitors like Codex, citing more consistent performance and better compute availability [1][6].

1. Google broke its promise to me – now ICE has my data (eff.org)

1705 points · 764 comments · by Brajeshwar

The Electronic Frontier Foundation has filed complaints with state attorneys general after Google allegedly broke its privacy promise by handing a student's data to ICE without prior notification, depriving him of the opportunity to challenge the administrative subpoena. [src]

The discussion highlights a growing distrust of Google, with some users citing this incident as their final motivation to migrate to self-hosted or privacy-focused alternatives like Proton Mail [0]. While some commenters question the specific legal details of the subpoena and whether Google technically violated its own non-disclosure policies [5], others argue that the core issue is the systemic weaponization of data by government agencies like ICE against individuals [3][9]. There is a strong consensus that such stories are vital for industry decision-makers to see, as they fundamentally alter the legal and ethical calculations of trusting major tech corporations with sensitive data [1][2].

2. Claude Design (anthropic.com)

1217 points · 750 comments · by meetpateltech

Anthropic has launched Claude Design, a new initiative from Anthropic Labs focused on exploring and sharing the design principles and creative processes behind the development of the Claude AI interface. [src]

The release of Claude Design has sparked a debate over whether AI-driven UI generation fosters efficiency or merely accelerates the "homogenization" of the web [0][6]. While some argue that standardized, "obvious" interfaces are ideal for functional tools like medical software, others contend that AI lacks the capacity for the original thought and "artisanal weirdness" required for truly groundbreaking design [1][8][9]. Critics warn that these tools may lead users to confuse output with agency, potentially blinding them to the deep structural problem-solving that defines professional design [2][4]. Conversely, proponents suggest that AI can accelerate learning by handling mundane tasks, allowing creators to focus on higher-level architecture rather than "tracking down stupid issues" [7][8].

3. Backblaze has stopped backing up OneDrive and Dropbox folders and maybe others (rareese.com)

1127 points · 690 comments · by rrreese

Backblaze has updated its backup client to automatically exclude folders from cloud storage providers like OneDrive, Dropbox, and Google Drive, as well as `.git` directories. Users are criticizing the company for implementing these exclusions silently without direct notification or clear documentation on their website. [src]

Backblaze's decision to exclude OneDrive and Dropbox folders from its personal backup service is seen by users as a breach of its "unlimited" storage promise and a failure to act as a reliable last-resort backup [0][3][8]. While some commenters suggest the change is a technical necessity to prevent "files on demand" features from crashing laptops by forcing massive downloads [1], others argue that excluding synced folders leaves users vulnerable to data loss if a sync service accidentally overwrites or corrupts files [3][5]. Critics contend that "unlimited" marketing is inherently unsustainable and signals that financial teams are prioritizing cost-cutting over data integrity [2][9].

4. Qwen3.6-35B-A3B: Agentic coding power, now open to all (qwen.ai)

1266 points · 531 comments · by cmitsakis

Alibaba has open-sourced Qwen3.6-35B-A3B, a sparse mixture-of-experts model with 3 billion active parameters that delivers high-performance agentic coding and multimodal reasoning. The model rivals much larger dense models and is now available via open weights, Qwen Studio, and the Alibaba Cloud API. [src]

The Qwen 3.6 release has sparked excitement for its agentic coding capabilities, with early users reporting it can outperform models like Opus 4.7 in specific creative tasks [2]. While there is relief that the Qwen team continues to publish open weights despite recent internal departures [3], some users expressed disappointment that the highly requested 27B variant was bypassed in favor of this 35B model [9]. Technical discussions focus on hardware requirements, noting that while 16GB GPUs may struggle with quality loss [1][7], quantized versions from providers like Unsloth allow the model to run on consumer laptops [0][2]. However, community members caution that launch-day quantizations often require later revisions to fix performance bugs [8].

5. Codex for almost everything (openai.com)

998 points · 554 comments · by mikeevans

OpenAI has released a major update to Codex, enabling the AI to operate computers alongside users, browse the web, generate images, and automate long-term developer workflows through new memory features and over 90 third-party plugins. [src]

The rise of "professional agents" like Codex and Claude Cowork is viewed by some as a potentially massive product category that could disrupt traditional software by allowing agents to interface with apps on behalf of non-technical users [2]. However, critics argue that these tools are merely catching up to existing features in Claude [3] and that non-technical users may find the unpredictable nature of AI-generated interfaces and "vague request" processing frustrating rather than helpful [7]. While some users find value in replacing CLI tasks with AI commands [9], others express significant security concerns regarding giving models direct control over their computers and applications [8]. There is also a cynical view that the current hype is driven by OpenAI's strategic use of subsidized compute to win a PR war against Anthropic [0][5][6].

6. Someone bought 30 WordPress plugins and planted a backdoor in all of them (anchor.host)

1194 points · 340 comments · by speckx

A malicious buyer acquired a portfolio of over 30 WordPress plugins and planted a sophisticated backdoor that remained dormant for eight months before injecting SEO spam via `wp-config.php`. WordPress.org has since closed the affected plugins, which include popular tools like Countdown Timer Ultimate and Popup Anything on Click. [src]

The incident highlights a critical vulnerability in modern software where attackers can simply purchase dependencies or bribe employees to insert backdoors, a tactic fueled by the massive financial incentives of cryptocurrency [0][7]. Commenters argue that the industry's reliance on massive trees of unvetted transitive dependencies makes supply chain attacks nearly inevitable [1][3][9]. While some debate whether "bug-free" software is even possible, others contend that we possess the technical tools to achieve high quality but consistently prioritize speed and cost over security [2][4][5][8].

7. The future of everything is lies, I guess: Where do we go from here? (aphyr.com)

728 points · 762 comments · by aphyr

Kyle Kingsbury argues that society should resist the adoption of large language models to preserve human skill and critical thinking, warning that AI's rapid integration threatens to cause profound cultural, economic, and psychological harm similar to the historical impact of the personal automobile. [src]

Commenters debate whether AI's societal impact will mirror the automobile, which some argue provided utility while causing deep cultural isolation and environmental harm [0][2][7]. While some fear AI will devalue human intellect and empower a small elite to control society [3][4], others contend the technology is currently too unreliable to replace human decision-making and is being overhyped to justify corporate layoffs [9]. Ultimately, there is a sense of unease regarding the shift in human values, as skills like writing and thinking may lose their status as primary drivers of upward mobility [4][5].

8. DaVinci Resolve – Photo (blackmagicdesign.com)

1145 points · 296 comments · by thebiblelover7

Blackmagic Design has introduced a dedicated Photo page to DaVinci Resolve, bringing its advanced Hollywood color grading tools, AI-powered effects, and RAW support to still photography. The update includes non-destructive editing, GPU-accelerated processing, and cloud-based collaboration for professional photographers and retouchers. [src]

Users are excited that DaVinci Resolve is bringing advanced video-centric color science and creative tools like relighting and film emulation to the stagnant photography market [0][3]. While some praise its performance on Linux via containerization [9], others report significant frustration with outdated audio APIs and codec support on the platform [2]. Early testers find the interface confusing and "tacked on" compared to Lightroom, suggesting that while the software is powerful, it currently lacks the intuitive workflow required to sway professional photographers [7].

9. IPv6 traffic crosses the 50% mark (google.com)

812 points · 614 comments · by Aaronmacaron

Google's tracking data shows that global IPv6 adoption has reached approximately 45.54%, reflecting the percentage of users who access the platform via the updated internet protocol. [src]

While IPv6 traffic has reached 50%, users observe a plateau in adoption driven by enterprise resistance and the protocol's inherent complexity [0][5]. Critics argue that IPv6 is not a simple expansion of IPv4 but a "recursive WTF" with unresolved issues regarding address selection, DHCP support, and fragmentation that break established operational practices [5]. Major platforms like GitHub remain IPv4-only, likely due to the risk of breaking customer IP-based access controls during a transition [1][8]. Consequently, many organizations continue to actively block IPv6 at the firewall, leading some to believe the protocol will never fully succeed in its current form [7][9].