Link to paper: https://
arxiv.org/pdf/2308.01074
@aibreakfast
-
AI for Realistic Human Face Generation with Style Control
By
–
-
AI model detects typing from keystroke sounds with 95% accuracy
By
–
There’s an AI model that can determine what you are typing just by listening to your keystrokes – with 95% accuracy:
-
ElevenLabs Launches Scribe V2 Realtime: The Best Speech-to-Text Model Yet
By
–
The champion has returned:
— AI Breakfast (@AiBreakfast) 9 janvier 2026
@elevenlabsio just released Scribe V2 Realtime, an ultra-accurate speech to text model with precision timestamps and entity detection, optimized for low-latency and voice agents.
Best transcription model we’ve seen yet. https://t.co/JgM7LEImmEThe champion has returned: @elevenlabs just released Scribe V2 Realtime, an ultra-accurate speech to text model with precision timestamps and entity detection, optimized for low-latency and voice agents. Best transcription model we’ve seen yet.
-
Normal LLMs store knowledge in high-precision weights causing huge memory
By
–
This is insane, here’s how this works:
— AI Breakfast (@AiBreakfast) 5 janvier 2026
Normal LLMs (like Grok or GPT variants) store their "kn0wledge" in billions of numbers (weights)
Each weight is usually a high-precision decimal (like 3.14159), stored using 16 bits per weight. This makes the model huge in memory and… https://t.co/P2W0pNNqxqThis is insane, here’s how this works: Normal LLMs (like Grok or GPT variants) store their "kn0wledge" in billions of numbers (weights) Each weight is usually a high-precision decimal (like 3.14159), stored using 16 bits per weight. This makes the model huge in memory and
-
Near-instantaneous open-source vision models demo with webcam
By
–
Near-instantaneous open-source vision models are here, and they’re cheap:
— AI Breakfast (@AiBreakfast) 4 janvier 2026
This is a real-time demo built with HuggingFace SmolVLM and llama.cpp using only a webcam
(Github link below) pic.twitter.com/yAkNt6P7PDNear-instantaneous open-source vision models are here, and they’re cheap: This is a real-time demo built with HuggingFace SmolVLM and llama.cpp using only a webcam (Github link below)
-
OpenAI’s pen-shaped AI device aims to be third core device
By
–
Additional details about OpenAI’s AI device: It is a pen-shaped device that integrates AI, aiming to become a “third core device” following the iPhone and MacBook. It’s lightweight and highly portable—about the size of an iPod Shuffle—and can be carried in a pocket or worn x.com/jukan05/status…
-
2026 AI Predictions: Code Generation Becomes Key Benchmark
By
–
Alright, here's my 2026 AI Predictions (What did I miss?): -All meaningful progress benchmarks shift to code gen capabilities. Models write looong systems, debug themselves, and ship real applications. Your dad will proudly show you an app he made. Short prompts expand into
-
DeepSeek rivals GPT-5 on LM Arena, released in August
By
–
For coding it is, but not for typical LLM outputs. DeepSeek is tied with GPT-5 on LM Arena and it was released in August.
-
Open-source AI on Mac Studios: Cheaper, faster, and free alternative to ChatGPT
By
–
Small companies could run DeepSeek-V3.2 on one of these souped-up Mac Studios instead of paying for dozens and dozens of ChatGPT licenses for employees. Open-source models are only about 6 months behind the proprietary models. And they’re free. Forever.
-
Easy AI business turning PDFs into JSON data for clients
By
–
If you ever wanted to start an EASY business with AI: Just find customers with a lot of paperwork (HOAs, municipalities, law firms) and use @doctlyai to convert their PDFs into useable JSON data so they or their customers can query it with a chatbot, and charge $500-5k/mo.