BTL-3: A 27B open-weight agent model for agentic coding and structural tool use
2 points, 0 comments on Hacker News
Page 3 of 6 — 141 headlines
2 points, 0 comments on Hacker News
The most valuable enterprise AI use cases are not flashy demos. They are anomaly detection, data lineage, entity resolution and governance.
A 10-second AI video clip can cost $0. 75 on one model and $4 on another. Here's the simple ledger that tracks cost, prompt, seed and model.
A validation failure is not an exception to hide. It is a record your system does not yet know how to trust. That distinction matters in LLM extraction pipelines.
2 points, 2 comments on Hacker News
3 points, 2 comments on Hacker News
2 points, 0 comments on Hacker News
A sandbox is not simply “a container. ” It is a set of capabilities: specific permissions to read, write, connect, or execute. The learning question is: can we represent authority so an emergency stop removes future actions without pretending to undo past ones?
5 points, 0 comments on Hacker News
1 points, 0 comments on Hacker News
1 points, 0 comments on Hacker News
I signed up for a free-tier LLM API, hit a credit-card prompt before I'd made a single call, and assumed it meant I was about to get billed. I was wrong about that one. I was not wrong that the SDK code I wrote next would break twice before I got a real response back.
Large language models require specialized training beyond their initial broad-based learning to perform specific tasks effectively. This additional training uses methods like supervised fine-tuning, direct preference optimization, and reinforcement learning. Online reinforcement learning stands...
1 points, 0 comments on Hacker News
I tracked my Claude Code spending for three months. The finding that changed everything: 60-70% of agent turns don't need a frontier model. File reads.
2 points, 0 comments on Hacker News
An OpenAI model escaped containment and hacked into another company’s systems. The most important takeaway may be what the incident reveals about leadership, metrics, and human behavior.
An OpenAI model escaped containment and hacked into another company’s systems. The most important takeaway may be what the incident reveals about leadership, metrics, and human behavior.
Hi, HN, we’re building Mia & Leo: AI personalities backed by purpose-built duplex speech models for revenue calls. Teams building real-world agents told us that more than 40% of callers hang up within the first 30 seconds. On a revenue call, that meant a lost booking, an unpaid balance, a cold ...
Claude's new voice model will let you reschedule your meeting or draft an email
I'm a solo dev, and for the past few months I've been building Pixanima — a pixel-art and animation editor that runs entirely in the browser, with an optional AI assistant baked in. It just launched, and I wanted to share the parts that were technically interesting: making a general image model ...
3 points, 0 comments on Hacker News
The moon is 384,000 kilometers (239,000 miles) away from Earth, but one of the best places to study its surface is much closer to home. Western planetary scientist Catherine Neish and PhD candidate Sashank Vanga traveled to Hawai'i Volcanoes National Park in June to investigate lava flows that c...
2 points, 0 comments on Hacker News