The wiki

Machine-learning researcher and generative-model coauthor whose work spans raw audio, continuous diffusion for language, and explanations of how image diffusion builds structure.

Founder of Learn Prompting and co-author of The Prompt Report and HackAPrompt, working on prompting methods, adversarial attacks, and reproducible evaluation.

Computer scientist and Discovery Loop cofounder whose work with Jeff Dean helped make distributed computing manageable for programmers, with contributions to MapReduce and TensorFlow.

Founder of Cline, whose work centers on coding agents, developer control and open-source tooling.

Investor in data and AI infrastructure whose work connects practitioner problems, research-driven startups, and the discipline of running better AI experiments.

Cerebras developer-experience leader whose technical writing and experiments emphasize verification, disciplined agent loops, and developer communities.

Founder of Conviction and host of No Priors, Guo invests in AI-native companies and argues that their lasting value depends on private context, trusted access and organizational change.

Founder of Seek AI, focused on using language models to turn business questions into data queries.

Sarah Sachs’s documented AI engineering work at Notion spans agent infrastructure, evaluation, distributed tool ownership and model procurement.

Andreessen Horowitz general partner investing in AI models, enterprise software, and infrastructure, with public theses on compute financing, integrated applications, and industrial robotics.

Machine learning researcher at Cursor whose work spans language-model research, open-source implementations, and the organization of the Conference on Language Modeling.

Microsoft’s chairman and CEO on enterprise AI, customer control of models, and turning capability into measurable value.

Microsoft’s chairman and CEO, whose AI leadership spans the OpenAI partnership, Copilot organization, and an emphasis on customer control through models, context, tools, and private evaluations.

Cognition co-founder Scott Wu’s work on Devin centers on AI that executes software tasks, learns from deployment, and leaves humans to decide what to build.

Founder of Linzumi and former OpenAI technical staff member whose work connects written specifications, model alignment, and coordination of AI coding agents.

Cerebras CTO and co-founder whose work includes enabling sparse-model training and advocating faster inference as a way to make reasoning models practical in real-time applications.

LLM research engineer and author whose books, code and technical essays explain language models through working implementations and practical experiments.

A conversational data-query startup whose enterprise data-agent expertise IBM planned to incorporate into watsonx AI Labs, as announced in June 2025.

Sarah Nagy’s case for natural-language analytics starts with interrupted analysts and waiting colleagues. The challenge is making generated queries dependable enough to guide business decisions.

Technical strategy, partnerships and community work focused on how frontier AI reaches practical use.

From camera-based board games to interactive segmentation, Roboflow’s story shows how better models change annotation, deployment, and the economics of computer vision.

Part 4 of 4 of Latent Space’s ICLR 2024 recap examines how pre-training on task data changes the comparison between transformers and recurrent models—and lets simpler architectures close the gap.

SemiAnalysis connects semiconductor manufacturing, AI infrastructure performance and operating economics through research, consulting, institutional models and open inference benchmarks.

Genesis Molecular AI CTO and former Meta research leader whose work spans Llama training, machine translation and distributed infrastructure.

Modal connects fast container startup, Python orchestration and GPU utilization to make custom cloud workloads easier to run—and economical to scale.

Browserbase combines hosted browsers, natural-language automation and human takeover to handle the state, access and reliability problems beneath an agent’s clicks.

A GPU infrastructure company that operates clusters and lets customers resell reserved capacity, developed from an audio AI lab's need to sublease its own compute.

Short experiments need flexible compute. Hardware financing needs long commitments. SF Compute connects them through resale, short reservations, and standardized delivery.

AI researcher whose collaborative work spans zero-shot language-model reasoning, self-training, differentiable categorical sampling and robot learning.

Founder of Lexica and builder of Debuild and Vectordash, whose work spans image generation, code generation, and browser agents.