Seattle Skeptics on AI

Tamara Weed, Apr, 30 2026

Stop wasting compute. Learn when to move past Chinchilla optimality and enter the overtraining regime to balance training costs with inference performance.

Tamara Weed, Apr, 29 2026

Learn how Chain-of-Thought prompting improves LLM reasoning by breaking complex problems into steps. Discover best practices, scaling secrets, and trade-offs.

Tamara Weed, Apr, 27 2026

Stop fighting your AI coding assistant. Learn how to use inline code context and context engineering to reduce revisions by 73% and master the art of vibe coding.

Tamara Weed, Apr, 26 2026

Learn how to create prompt style guides to ensure AI-generated code remains consistent across sessions, reducing technical debt and review time.

Tamara Weed, Apr, 25 2026

Learn how to balance retrieval precision and computational cost by choosing the right embedding dimensionality for your LLM RAG system.

Tamara Weed, Apr, 24 2026

Explore how multimodal generative AI is ending the 'accessibility gap' through adaptive interfaces, real-time narration, and conversational descriptions for all users.

Tamara Weed, Apr, 23 2026

Learn how Large Language Models use token prediction and probability distributions to generate text, from the softmax function to decoding strategies like Top-P and Temperature.

Tamara Weed, Apr, 22 2026

Learn how to select the best embedding models for your enterprise RAG pipelines. Compare BGE-M3, OpenAI, and NVIDIA models to optimize accuracy and latency.

Tamara Weed, Apr, 21 2026

Explore how Multimodal LLMs are replacing traditional OCR with visual reasoning to extract complex tables, handwritten notes, and structured data from documents.

Tamara Weed, Apr, 20 2026

Explore how Multimodal Large Language Models (LAMs) are revolutionizing audio understanding, from spectrogram processing to real-time voice reasoning.

Tamara Weed, Apr, 19 2026

Explore how vibe coding is transforming software engineering in 2025, shifting the role from manual coding to AI orchestration and high-level system architecture.

Tamara Weed, Apr, 18 2026

Learn how to manage petabyte-scale LLM datasets using sharding, tiered storage, and sharded data parallelism to eliminate GPU idling and memory errors.