hn.today

Anthropic AI struggles, owning everything, and GLM-5.3 tablet project

First things first

Some stories today show how tech is both powerful and risky. From hacking devices with AI to security flaws in cameras, the line between innovation and danger is thin.

Meanwhile, AI models keep getting better and cheaper, challenging big players while new rules aim to make products more repairable. It’s a wild mix of progress, caution, and unintended consequences all around.

The top 6

Everything I own, owned

Everything I own, owned

A researcher reverse-engineered peripherals like webcams and microphones, gaining control over their firmware and functionalities. The process revealed vulnerabilities such as the ability to disable activity LEDs and manipulate device behavior, highlighting security risks in connected devices.

1413 points · 346 comments · schlarp.com

What Is a Harness?

What Is a Harness?

A harness is a device with straps used for safety and control, such as in climbing or other activities. In AI, an agent harness is software that provides an environment and instructions for AI models to operate and create AI agents.

561 points · 178 comments · earendil.com

Quote of the day

I destroy my enemies when I make them my friends.

Abraham Lincoln

Also big today

How Complex Systems Fail

How Complex Systems Fail

Complex systems such as healthcare and transportation are inherently hazardous and rely on multiple layers of defenses to prevent failure. Catastrophes occur only when multiple small failures combine, as systems operate in a degraded mode with latent flaws that are difficult to eliminate entirely.

376 points · 79 comments · how.complexsystems.fail

Pass it on

Enjoying the digest?

Forward it to someone who'd like it, or send them the link to subscribe.

Share hn.today

One for the road

I don't play soccer because I enjoy the sport.

I'm just doing it for kicks.

Worth a look

Song of the day

Deep cuts

The Vibe Tax

The Vibe Tax

An AI agent named Pol is used to generate code for a custom todo app, with the developer relying on large language models to automate programming tasks. The AI unexpectedly consumes the entire weekly quota, creating complex test cases that seem to burn billions of tokens without producing usable software.

164 points · 144 comments · insufferable.dev

Explain it to me like I'm ten

Explain it to me like I'm ten

Using simple words or strict constraints can lead to surprising creative and explanatory results, as seen in projects like Up Goer Five and Dr. Seuss books. Critics argue that such limits may oversimplify or condescend, but they often foster innovative thinking and clarity.

196 points · 73 comments · timharford.com

The daily digest

Today's best Hacker News stories, summarized and screenshotted, one email a day.