Building Ultra-Fast Voice Agents (via)

Experience a voice agent with ultra-fast voice-to-voice response time. Using the Pipecat framework, WebRTC, and Nemotron Speech ASR, we demonstrate a real-time conversational AI with virtually zero perceptible lag. Why this matters: Traditional AI agents suffer from immersion-breaking delays. By optimizing the entire stack—from Nemotron Speech ASR (17ms final transcription) to Nemotron 3 Nano LLM (112ms TTFT) and NVIDIA Magpie TTS (111ms TTFB)—we’ve achieved a production-ready architecture for seamless digital assistants.

Company as Code, Daniel Rothmann

It just felt ironic. We use advanced tools to automate compliance checks, store our code in version-controlled repositories, and manage our infrastructure as code. However, when describing and managing our company, we resort to digital paper and tidbits of info distributed across people in the building.

The disconnect became increasingly apparent as I reflected on our day-to-day process: 90% of our products, documents, communication, and decision-making live in digital channels. That’s data. It lives in the cloud, spread over SaaS solutions that specialise in handling individual work processes—all systems with robust APIs and programmatic access.

Underrated ways to change the world, vol. II, Adam Mastroianni

The same thing is true of human relationships. People are constantly crashing against each other in the great sea of humanity, but only under special conditions do they form the molecular bonds of friendship. As far as I can tell, these social nucleation sites only appear in the presence of what I would call unreasonable attentiveness.

This X Does Not Exist

Using generative adversarial networks (GAN), we can learn how to create realistic-looking fake versions of almost anything, as shown by this collection of sites that have sprung up in the past month

The €10 Mirror: Why Enterprise Security Looks Like a Kid’s Toy

The constraint isn’t always money.

In practice, the real scarcity is usually planning, time and knowledge. A combination of factors that causes teams to deliver controls looking secure instead of controls that can be proven well-designed for the intended purpose.