DevPick

News (page 16)

A report on using GLM 5.3 Flash for coding for one month detailed 2 billion tokens of usage and factors affecting model selection and costs.

  • •There was a total of 2 billion tokens of AI usage during one month of coding.
  • •The report analyzed elements influencing model selection criteria and cost structures.
#Generative AI Tools & Use#LLM / Model Trends
10/3/2026

An article explains how developers should strengthen AI instruction, critical review of outputs, and retention of technical judgment as AI agent adoption grows.

  • •The importance of the ability to effectively instruct AI agents is highlighted.
  • •Skills for critically reviewing and verifying AI-generated outputs are required.
  • •Placing technical judgment at the center of workflows is emphasized.
#Generative AI Tools & Use#AI Agents
10/3/2026

Unread: Claude-Shaped Science(opens in a new tab)

Hacker NewsHN submitter: famouswaffles
Sign up to see

Matthew Schwartz discussed changing his problem-solving approach with Claude to develop BootLoops, a rigorous calculation toolkit for quantitative science.

  • •He shifted from opposing Claude to finding 'Claude-shaped' problems that Claude excels at.
  • •He built a toolkit called BootLoops for rigorous calculations in quantitative science.
  • •The toolkit is being applied across multiple scientific fields in collaboration with experts.
#Generative AI Tools & Use#Research & Papers
10/2/2026

DeepSeek Harness desktop version released, enabling document processing, spreadsheet analysis, code creation, and task scheduling.

  • •DeepSeek Harness is provided as a desktop application for macOS and Windows.
  • •It supports interacting with documents, analyzing spreadsheets, creating code, and scheduling tasks.
  • •It features a mechanism to extend functionality through composable plugins.
#AI Agents#Generative AI Tools & Use
10/2/2026

An open-source identity and messaging system enabling stable communication between AI agents across sessions, runtimes, machines, and organizations.

  • •Provides open-source identity management features for AI agents.
  • •Features persistent messaging capabilities across sessions and machines.
  • •Supports wake-up events between AI agents across organizations.
#AI Agents#Generative AI Tools & Use
10/2/2026

Janus, an API router for AI models written in Go, was released, enabling the execution of GGUF models via Vulkan.

  • •Janus is an API router for AI models written in Go.
  • •It features a Vulkan model runner and operates on AMD, Intel, and Nvidia GPUs.
  • •It is provided as a binary for executing model files in GGUF format.
#Generative AI Tools & Use#Semiconductors & Infrastructure
10/2/2026

The Guided Vision feature is being integrated into Gemini Live, providing real-time audio descriptions and small text reading assistance for objects viewed through the camera.

  • •The feature is being introduced to Gemini Live on supported Android devices today.
  • •By sharing the camera, the AI provides audio descriptions of the surrounding environment.
  • •It assists with reading small text and identifying or finding objects in the vicinity.
#Generative AI Tools & Use#Major Players
10/2/2026

OpenAI added a shopping feature to ChatGPT allowing users to virtually try on clothing and accessories using their own photos and save favorites.

  • •OpenAI rolled out a new shopping feature in ChatGPT, enabling users to virtually try on clothing and accessories using their own photos.
  • •Users can save items they like to a 'favorites' library.
  • •The feature was released to expand the shopping experience within ChatGPT.
#Generative AI Tools & Use#Major Players#Business Use Cases
10/2/2026

An article analyzes AI-specific expression habits in Opus 5.5 outputs, highlighting differences from human writing.

  • •A notable feature of Opus 5.5 outputs is that the word 'dependable' appears 23 times more frequently than in human samples.
  • •It mentions 'tells' for distinguishing AI-generated text from human writing.
  • •The analysis focuses on specific vocabulary usage biases in AI-generated text.
#LLM / Model Trends#Generative AI Tools & Use
10/2/2026

Two BM25 optimizations inspired by PlanetScale's TIN benchmark and benchmark configuration changes sped up ParadeDB's text search without altering document identifiers.

  • •Two BM25 optimizations inspired by PlanetScale's TIN benchmark were implemented.
  • •ParadeDB's text search was sped up, accompanied by changes to benchmark configurations.
  • •Performance improvements were achieved without changing document identifiers.
#Semiconductors & Infrastructure#Generative AI Tools & Use
10/2/2026