Google Lança Modelos Gemma 4 e Muda para Licença Apache 2.0

Google introduced the Gemma 4 family of open-weight AI models, offering four variants optimized for local execution and mobile devices. The two larger models—26B Mixture of Experts and 31B Dense—run unquantized on a single 80GB Nvidia H100 GPU and can be quantized for consumer GPUs. Smaller Effective 2B and Effective 4B models target smartphones and edge hardware, benefitting from collaboration with Qualcomm and MediaTek. Google also replaced its custom Gemma license with the Apache 2.0 license, giving developers greater freedom. The company claims Gemma 4 models are the most capable locally runnable AI systems, positioning them near the top of open AI model rankings. Ler mais

Jan 26, 2026

Dispositivo DIY ‘Thought Catcher’ permite gravar ideias à noite sem um telefone

A maker has turned a Raspberry Pi into a screenless device called the Thought Catcher that sits by the bedside. Press a button, speak your idea, and the AI‑driven Whisper service transcribes the audio and uploads it to Notion. The system adds a beep to confirm completion, letting users capture thoughts without reaching for a phone and risking sleep disruption. The device requires a Raspberry Pi, a button, a microphone, a buzzer and custom Python code. Future versions may be refined, and the creator plans to share build instructions if interest grows. Ler mais