Category: AI
-
Essay on hallucination and multiple methods to eliminate it
Essay on hallucination and multiple methods to eliminate it https://lilianweng.github.io/posts/2024-07-07-hallucination
-
Claude Update
Key points The post reports that Claude is updated to do thinking and be less overactive. The article describes the update as possibly in preparation for a Lex Fridman podcast. Claude is updated to do thinking and be less overactive, possibly in preparation for Lex Fridman podcast.
-
Janus from DeepSeek Completes the Cycle
Key points According to the post, Janus from DeepSeek can perform normal text generation with image understanding as well as create images. Janus from DeepSeek can perform normal text generation with image understanding as well as create images.
-
Suno released a feature that creates song ideas from images
Key points Suno released a feature that creates song ideas from images. Suno released a feature that creates song ideas from images
-
Mistral Introduces Small Models for Edge Devices
Key points Mistral has released small but capable models intended for usage on edge devices. Mistral has released small but capable models intended for usage on edge devices.
-
Canvas, Open AI’s response to Antrophic Artifacts
Key points Canvas is available in beta for ChatGPT Plus users. The post describes Canvas as OpenAI’s response to Anthropic Artifacts. Canvas, Open AI’s response to Antrophic Artifacts, is available in Beta for ChatGPT Plus users
-
Advanced Voice Mode API
Key points OpenAI launched the Advanced Voice Mode API. The API enables developers to build more sophisticated voice applications. OpenAI launched the Advanced Voice Mode API, enabling developers to build more sophisticated voice applications.
-
Meta AI Can Mimic Voices
Key points Meta AI can now speak in the voices of Awkwafina and John Cena, among others. Meta AI can now talk in the voice of Awkwafina and John Cena, among others.
-
Advanced Voice Rolling Out to ChatGPT Plus and Team Users
Key points OpenAI said Advanced Voice was rolling out to all Plus and Team users in the ChatGPT app over the course of that week. It includes five new voices, improved accents, and the ability to say “Sorry I’m late” in over 50 languages. Advanced Voice is rolling out to all Plus and Team users […]
-
Bytedance Releases PixelDance & Seaweed Video Generation Models
Key points Bytedance released 豆包PixelDance and Seaweed, two video generation models. Bytedance has just released 豆包PixelDance & Seaweed, video generation models. However, there is no English cover yet.