Beam 1.0 is a chat-template and system specification for an AI assistant named Beam, encoded as a Jinja template that defines identity, runtime knobs, tool conventions, and a wire format for tool calls. It prescribes configurable kwargs (model_identity, reasoning_effort, builtin_tools, allow_missing_thinking), maps human-readable reasoning labels to numeric effort budgets (minimal=8 ... max=512), and enforces rules for builtin tools (only specific search/browse definitions allowed, a special-case bash). It defines how OpenAI roles map into blocks, how consecutive system/developer messages join, and a compact tool wire protocol (XML-like calls with analysis/results semantics and function.arguments as dicts).
The template's substantive logic is a suite of macros to convert JSON Schema into TypeScript types and to render tool declarations and call bodies. Key macros include json_schema_to_typescript, render_nullable_type, render_default, and render_tool_declaration; they handle enums, oneOf variants, objects (with ordered properties, required flags, descriptions, examples, defaults), arrays, and primitive type mapping (integer→number). Tool declarations output either a simple string header for pure-string parameters or a TypeScript function type for structured parameters; render_json_value and render_parameter_value control JSON encoding. Overall, the template automates generation of typed tool interfaces and prompt wiring for reliable tool calls and developer-facing documentation.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Open weight models now process over 71.5% of tokens and account for nearly 23% of AI Gateway spend, quadrupling their share since June. Spending on open weight models has increased more than four times in three months, reflecting rapid adoption and shifting AI usage patterns. (vercel.com)
An AI system has successfully designed 16 new viruses, including bacteriophages. This breakthrough raises concerns about potential biosecurity risks. (cnn.com)
Flux 3 Image Klein is a 5.6 billion parameter open weights model for text-to-image generation, editing, and multi-reference creation. It offers high-quality, ground-connected images with efficient performance on consumer GPUs, enabling local and on-device use. (bfl.ai)
Yum! Brands has implemented AI in its restaurants, such as automating drive-thru orders, which has improved employee retention by making jobs easier. Despite public criticism of AI, company leaders see its benefits in increasing workplace satisfaction and efficiency. (fortune.com)
Transformers function as sophisticated memory systems that recall information via self-attention, minimizing interference and boosting capacity. Different neural layers, from linear maps to self-attention, are explained as various forms of memory storage and retrieval. (alexwang.ai)
Granite Switch 4.1 integrates chat, decision-making, and speech-to-text functions into a single endpoint, enhancing efficiency and speed. It features models optimized for safety, decision tasks, and transcription, all with reduced GPU requirements. (huggingface.co)
Today's best Hacker News stories, summarized and screenshotted, one email a day.