THE DECODER
DeepL launches new AI model for translation and text editing that supposedly beats GPT-4
Technology
Economy
DeepL unveils new AI model surpassing GPT-4; offers improved translations in English, Japanese, German, and Simplified Chinese.
2024-07-18 19:55:08
Artificial Intelligence
Technology
Business
Advancements in AI Models: DeepL, OpenAI, and Google DeepMind
DeepL introduces a new AI model that exceeds GPT-4 in translation capabilities. OpenAI launches GPT-4o Mini, a cheaper and smarter model offering higher performance.
Google DeepMind develops an open-source AI system called FLAMe to reduce biases in language model evaluation. These advancements in AI models aim to provide improved translations, cost-effective options, and unbiased evaluations of text quality.
OpenAI's ChatGPT, a popular AI-powered chatbot, gains widespread adoption and partnerships with companies like Apple and Reddit. However, OpenAI faces challenges such as internal conflicts, copyright infringement lawsuits, and concerns regarding AI ethics.
Despite these challenges, OpenAI continues to expand its offerings, with plans for a GPT-5 model and partnerships with major publishers like The New York Times and Financial Times. The development of AI models like ChatGPT raises questions about privacy, misinformation, and the responsible use of AI technology.
Related news on that topic:
The press radar on this topic:
THE DECODER
Google Deepmind develops open-source AI to tackle biases in evaluating language models
Technology
FLAMe, an open-source AI system developed by Google DeepMind, Google, and UMass Amherst, outperforms GPT-4 in evaluating text quality, with lower bias. FLAMe-RM variant achieves 87.8% accuracy on RewardBench. Researchers aim to enable open and transparent evaluation, using data from publicly available sources, and plan to release the training data and models. However, they caution that excessive use of such systems may amplify existing biases in the training data.
The Verge
OpenAI is releasing a cheaper, smarter model
Technology
Finance
OpenAI is releasing a cheaper, lighter model called GPT-4o Mini, which scores 82% on the MMLU benchmark and is more capable than GPT-3.5. Developers like Ramp and Superhuman have tested the model, using it to build expense data extraction and email auto-suggestion features. The new model will be available to ChatGPT users, replacing GPT-3.5 Turbo, providing a more affordable option for AI applications.
TechCrunch
OpenAI unveils GPT-4o mini, a small AI model powering ChatGPT
Technology
Economy
GPT-4o mini has a context window of 128,000 tokens and a knowledge cutoff of October 2023, making it a cost-effective AI model for developers.
Welcome!
infobud.ai is an AI-driven news aggregator that simplifies global news, offering customizable feeds in all languages for tailored insights into tech, finance, politics, and more. It provides precise, relevant news updates, overcoming conventional search tool limitations. Due to the diversity of news sources, it provides precise and relevant news updates, focusing entirely on the facts without influencing opinion. Read moreExpand