
Benedict Evans Says Most Enterprise AI Spend Is Wasted — Here's What Actually Works
Benedict Evans' September 2026 essay landed on Hacker News with 153 upvotes and a blunt claim: compa...
Deep dives on the models, agents and automation actually worth your time — what they do well, where they break, and what it costs to run them.
33 articles published

Benedict Evans' September 2026 essay landed on Hacker News with 153 upvotes and a blunt claim: compa...

A study of Google's AI Mode found it surfaces products averaging 21.6% more expensive than regular s...

A viral Dan Luu post put AI skeptic Ed Zitron's two years of bearish predictions under a microscope,...

The EFF is asking federal courts to reject a legal theory that would give AI companies a fair-use ex...

Apple's own supply chain didn't see this coming: AI demand has turned the Mac Mini and Mac Studio, t...

Anthropic pays a model to quit conversations it finds distressing, while most neuroscientists say th...

Nvidia says it will book $673 billion in AI-related sales, a number so large it forces a reckoning w...

Apple just put local AI silicon into its highest-end Macs, and the move quietly undercuts the Nvidia...

Anthropic built the model researchers rank highest, and it still can't out-earn cheaper competitors ...

LinkedIn just gave users a button to flag posts as AI slop, and the company's own detection systems ...

Wiz researchers traced a compromise of Snowflake's Jira instance back to a single AI generated code ...

404 Media physically tracked a shipment of scarce, out-of-print books to an Amazon facility tied to ...

A single prompt sent to 11 different AI models produced 11 genuinely different answers, exposing how...

AI agents promise to book your flights, manage your inbox, and run your errands autonomously — but a...

Everyone's chasing AI hype, but the real money is flowing into power grids, data centers, and chip f...

Every AI product hits the same wall: inference is slow, GPUs are expensive, and nobody explains why....

A viral Hacker News essay just upended the AI-deskilling narrative with a simple claim: LLMs reward ...

A single Hacker News thread with over 900 upvotes just upended assumptions about who leads in AI cod...

GitHub just shipped native stacked pull requests, a workflow that's been the domain of third-party t...

A single link posted to Hacker News racked up 360 upvotes by revealing something unsettling: the web...

A quiet browser-based tool just resurrected the software philosophy that shaped early Mac computing,...

HackerRank just open-sourced the AI model that scores your resume, and the results are chaotic. We d...

A little-known model named GLM 5.2 just outperformed Anthropic's flagship in rigorous cybersecurity ...

A quiet podcast from a West Texas radio station is trending as the ultimate sleep aid. This isn't ju...

The brutal economics of LLM inference are a bottleneck on innovation. A new paper from DeepSeek-AI i...

The xz-utils backdoor wasn't a bug; it was a market failure years in the making. We’re now facing a ...

Apple's shocking 2026 price hikes on MacBooks and iPads aren't just a simple adjustment; they're the...

OpenAI just dropped a bombshell, revealing its first custom AI chip in partnership with Broadcom. Th...

Programming legend John Carmack recently revealed his biggest early career mistakes. We go beyond th...

The constant flood of CVEs and security alerts is no longer a feature, but a bug in the system. We'r...

Meta just tipped its hand: a clip-on AI pendant that remembers your whole day, four new smart glasse...

Nvidia CEO Jensen Huang has committed $150 billion to Taiwan's AI infrastructure — the largest corpo...

Google redrew the developer stack at I/O 2026. Antigravity 2.0 is no longer a coding assistant — it'...