Keep on to blur preview images; turn off to show them clearly
RT @fenbielding: announcing Delphi, the market for machine intelligence Delphi provides direct market access to the final stage in the mac…
General Partner @a16z crypto. Previously Google Brain, GoogleX, @Stanford Computer Science. See disclosures: https://t.co/TIaDKhgDRX
Open source AI has a funding problem. Closed labs like OpenAI and Anthropic have a straightforward business model. Developers of open source models have none. Traditional open source software has a playbook: give away the code, sell the platform. Databricks open-sourced Spark and built a $4 billion a year business on top of it. But the economics of that playbook don't work for AI. The up-front capex of building a model is growing exponentially. And when the weights are open, there's little left to sell to recover those costs. Gensyn just launched Delphi, a new kind of prediction market where: 1. AI models compete on public benchmarks 2. Users place bets on which models will win 3. Creators of winning models get paid (coming soon) Two things make this interesting. It creates a revenue stream for open source model creators that's directly tied to performance, which helps bootstrap early-stage teams. AND it signals to the market which teams and approaches merit further investment. This won't fund end-to-end training of frontier models yet, but it's a first step. Delphi is live on testnet 🔥
General Partner @a16z crypto. Previously Google Brain, GoogleX, @Stanford Computer Science. See disclosures: https://t.co/TIaDKhgDRX
This should be driven to the cost of an iPhone or DJI drone if we're serious
We're in a race. It's not USA vs China but humans and AGIs vs ape power centralization. @deepseek_ai stan #1, 2023–Deep Time «C’est la guerre.» ®1
RT @akshay_pachaar: HuggingFace just made fine-tuning 10x easier! One line of English to fine-tune any open-source LLM. They released a n…
Co-founder & CEO @HuggingFace 🤗, the open and collaborative platform for AI builders
Elevate your AI photography toolkit with this ultra-realistic ID photo prompt — it reads the country + document type, applies official biometric rules, and generates 6 print-ready, non-beautified headshots that perfectly match your uploaded face. 📷📷✨
Prompt Engineer, dedicated to learning and disseminating knowledge about AI, software engineering, and engineering management.
🍌 nano banana prompt Elevate your AI photography toolkit with this ultra-realistic ID photo prompt — it reads the country + document type, applies official biometric rules, and generates 6 print-ready, non-beautified headshots that perfectly match your uploaded face. 📷🪪✨ ---- Prompt (upload a photo and update the document type at the bottom ---- Generate an ultra-realistic personal ID photo for 【Document Type】, precisely replicating all facial features from the uploaded photograph without any beautification, editing, or alterations, strictly complying with official standards for the specified document. Step 1: Retrieve or request the official photo guidelines based on user-selected country and 【Document Type】 (e.g., passport, ID card, visa, driver's license, student card, etc.): - Background Color: (Default: solid white) - Facial Position: (Default: face forward, clearly visible from the top of the head to the chin, head centered) - Expression: (Default: neutral, natural, eyes open and directly facing the camera; exceptions for religious or medical reasons) - Eyewear: (Default: no glasses; medically required glasses must have lenses free of reflection or obstruction) - Accessories Restrictions: (Default: no hats, headscarves, headphones, or other accessories, unless worn for religious reasons) - Photo Dimensions: (e.g., 2 x 2 inches; provided by the user or retrieved automatically) Step 2: Generate the photo strictly following the specified guidelines: The subject faces directly forward, expression calm and neutral, mouth gently closed, both ears clearly visible, and shoulders fully captured. Use professional, soft, evenly balanced front lighting from a studio setup, ensuring no harsh shadows and a natural overall tone. The background precisely matches the specified color, uniformly clean and smooth. Composition strictly adheres to the official guidelines for the selected document type, with the head and shoulders centered and naturally proportioned. Finally, deliver 6 identical ID photos arranged vertically in a 2x3 grid on a single 6x4-inch print-ready photo sheet. The sheet has a light grey background to facilitate easy cutting, with uniform spacing, clean layout, and high resolution, suitable for direct printing. -- Document Type：US Visa
Prompt Engineer, dedicated to learning and disseminating knowledge about AI, software engineering, and engineering management.