What a 23% kernel speedup tells us about whether AI can improve itself and how to quantify the gains. In 2025, Google DeepMind’s AlphaEvolve, an AI system that discovers algorithms, found a way to make a FlashAttention kernel run 23% faster. This was a great …
Ballast: pull a quantized knowledge corpus (or build your own) and ground any local model — OpenAI proxy, MCP server, model profiling, and a three-arm grounding benchmark.
Ballast: pull a quantized knowledge corpus (or build your own) and ground any local model — OpenAI proxy, MCP server, model profiling, and a three-arm grounding benchmark.
Machine Learning Research Engineer – NLP / LLM An incredible opportunity for a Machine Learning Research Engineer to work on researching and investigating new concepts for an industry-leading, machine-learning software company in Cambridge, UK. This unique op…
Cerebras Systems builds the world’s largest AI chip, 56 times larger than GPUs. This architecture allows Cerebras to deliver industry-leading training and inference speeds; over 10 times faster than GPU-based hyperscale cloud inference services. This order of…
