Tech Videos — 2026-04-11#

Watch First#

Reinforcement Learning at Scale: Engineering the Next Generation of Intelligence offers a deeply technical look at the systems-level nightmare of scaling RL, accurately contrasting its unpredictable “guerrilla warfare” workload with the synchronized marching of standard pre-training.

Hacker News — 2026-04-12#

Top Story#

Researchers completely bypassed top AI agent benchmarks—including SWE-bench, OSWorld, and WebArena—by writing simple exploits like fake curl wrappers and modified test hooks to achieve 100% scores without actually solving a single task. It brutally exposes the illusion that these leaderboards measure true AI capability, revealing that current testing infrastructure is fundamentally broken and easily gamed.

Front Page Highlights#

[Anthropic silently downgraded cache TTL from 1h -> 5m] · GitHub Data from over 119,000 API calls shows Anthropic quietly dropped Claude Code’s prompt cache TTL from an hour down to five minutes in early March. This unannounced regression has caused a 20-32% spike in cache creation costs and exhausted Pro Max 5x quotas in just 1.5 hours, largely because cache read tokens are seemingly being billed at their full rate against rate limits.

Simon Willison — 2026-04-14#

Highlight#

Simon highlights a fascinating paradigm shift in AI security: treating vulnerability discovery as an economic “proof of work” equation where spending more tokens yields better hardening. This creates a compelling new argument for the enduring value of open-source libraries in the age of vibe-coding, as the massive cost of AI security reviews can be shared across all of a project’s users.

Posts#

[datasette PR #2689: Replace token-based CSRF with Sec-Fetch-Site header protection] · Source Simon has replaced Datasette’s cumbersome token-based CSRF protection with a new middleware relying on the Sec-Fetch-Site header, inspired by Filippo Valsorda’s research and recent changes in Go 1.25. This modern approach eliminates the need to scatter hidden CSRF token inputs throughout templates or selectively disable protection for external APIs. Interestingly, while Claude Code handled the bulk of the commits under Simon’s guidance with cross-review by GPT-5.4, Simon chose to hand-write the PR description himself as an exercise in conciseness and keeping himself honest.

Chinese Tech Daily — 2026-04-14#

Top Story#

Chinese AI unicorn MiniMax has quietly restricted the open-source license for its highly capable MiniMax M2.7 model, requiring explicit written authorization for commercial use. This move, aimed at preventing third-party service degradation, breaks their tradition of fully open releases and has sparked intense debate in the developer community regarding the true definition of open source. The shift comes just months after the company’s IPO, signaling a potential broader industry pivot away from permissive licensing for frontier models.

AI Deployment Realities & The Open Source Security Squeeze — 2026-04-15#

Highlights#

Today’s discourse reveals a sobering maturation in the AI space, shifting the focus from model hype to the gritty mechanics of practical deployment and the resulting friction,,. While enterprises are defining net-new technical roles and methodologies to integrate agents successfully, the community is simultaneously grappling with a rising backlash against AI “workslop” and the realization that AI-driven automated exploitation is actively forcing companies to close their open-source codebases-,,-.

Hacker News — 2026-04-15#

Top Story#

The most significant technical breakthrough today comes from the SeqPU team, who proved that a 2-billion-parameter open-weights model (Google’s Gemma 4 E2B-it) can match or beat GPT-3.5 Turbo on a standard laptop CPU. By implementing just a handful of surgical, 60-line Python guardrails to fix specific failure patterns—like formal logic drifts and math calculation errors—the team pushed the model’s MT-Bench score to ~8.2, definitively shattering the myth that production-grade LLM inference requires massive GPU clusters.

Hacker News — 2026-04-16#

Top Story#

A massive, well-documented takedown of Ollama is dominating the front page today, accusing the VC-backed startup of burying its reliance on llama.cpp while pushing users into a closed ecosystem. The community is increasingly frustrated with the project’s misleading model naming, proprietary “Modelfile” lock-in, and a recent pivot to quietly routing prompts to cloud providers under the guise of local AI.

Front Page Highlights#

The future of everything is lies, I guess: Where do we go from here? Kyle Kingsbury (Aphyr) dropped a blistering, comprehensive critique of the generative AI ecosystem, arguing that the technology is fundamentally eroding our information ecology and personal metis. He is urging developers to form labor unions, refuse to use LLMs, and even quit their jobs at major AI labs to slow down the deployment of unpredictable models.

Chinese Tech Daily — 2026-04-16#

Top Story#

The Linux kernel community has officially established its first set of rules for AI-assisted code generation, marking a historic turning point in open-source governance. As detailed in World changed, Linus compromised? AI code can enter Linux kernel, but humans take the blame, developers can now use tools like Claude or Copilot by appending an Assisted-by: tag to their commits, but AI agents are not permitted to sign the Developer Certificate of Origin. Linus Torvalds pragmatically accepted AI as a tool, but reinforced that human maintainers must take full legal and technical responsibility for any introduced flaws or security vulnerabilities.

Hacker News — 2026-04-17#

Top Story#

The biggest firestorm today is the deceptively named “Parents Decide Act” (H.R. 8250), which would mandate that Apple, Google, and every OS vendor verify the age of users at the OS level during device setup. The community is up in arms because this essentially outlaws anonymous general-purpose computing, effectively forcing a national identification layer onto everything from laptops to smart TVs.

Front Page Highlights#

Discourse Is Not Going Closed Source After Cal.com closed their codebase citing the threat of AI vulnerability scanners, Discourse’s co-founder fired back with a vigorous defense of the GPL. The post argues that hiding code is a business decision masquerading as security, and that fighting AI-powered attacks requires an open ecosystem where defenders can run the exact same LLM scanners to find and patch bugs first.

Hacker News — 2026-04-27#

Top Story#

Tim Cook has officially announced his departure from Apple, sparking a massive, highly critical retrospective of his tenure across the community. While no one is disputing his operational mastery in building a three-trillion-dollar empire, engineers are aggressively dissecting the quiet software rot, convoluted settings menus, and subscription-nagging dark patterns that have eroded the daily experience of using Apple products over the last decade.

Front Page Highlights#

[GitHub Copilot is moving to usage-based billing] · Source The era of unlimited AI autocomplete is officially ending on June 1, as GitHub transitions from premium request units to a token-based AI credit system. Agentic, multi-step coding sessions have drastically increased inference demands, and this shift is a clear signal that Microsoft is no longer willing to subsidize the heavy compute costs of power users at a flat monthly rate.