According to @karpathy, the nanochat d32 language model was recently taught to count occurrences of the letter 'r' in words like 'strawberry' using a new synthetic task called SpellingBee (source: github.com/karpathy/nanochat/discussions/164). This process involved generating diverse user queries and ideal assistant responses, then applying supervised fine-tuning (SFT) and reinforcement learning (RL) to instill this capability in the AI. Special attention was given to model-specific challenges such as prompt diversity, tokenization, and reasoning breakdown, especially for small models. The guide demonstrates how practical skills can be incrementally added to lightweight LLMs, highlighting opportunities for rapid capability expansion and custom task training in compact AI systems (source: @karpathy on Twitter).

According to KREA AI (@krea_ai), the newly launched Krea 1 image generation model is designed to maximize artistic exploration and avoid the common 'AI look' typically seen in other image models. Krea 1 enables the creation of images with detailed skin textures, dramatic camera angles, cinematic lighting, and hundreds of expressive styles. The platform also supports style references and custom training, providing artists and creative businesses with greater flexibility and control over visual output. This development opens up significant opportunities for AI-powered content creation in advertising, gaming, digital art, and media industries, setting a new standard for photorealism and creative expression in generative AI models (Source: @krea_ai, June 17, 2025).