Agile: AI Updates - Week of [07/29/24]

The latest AI newsletter updates every week.

Overview - Meta’s Llama 3.1 model introduces advanced multilingual and reasoning capabilities to AI. Mistral and OpenAI are also pushing boundaries with Mistral’s new model and SearchGPT, while ethical discussions around AI privacy and data usage grow.

0:00

1. Meta Releases Llama 3.1 405B Model

• Meta has launched the Llama 3.1 405B model, a state-of-the-art open-source large language AI. This model boasts advanced multilingual capabilities, enhanced reasoning, and sophisticated tool-use features, marking a significant leap in AI technology.

2. Mistral's New Large Language Model

• Mistral has introduced the Mistral Large 2, featuring 123 billion parameters and a 128k context window. This model supports multiple languages and programming languages, offering extensive applications in various fields.

3. AI Solves Mathematical Olympiad Problems

• AI has achieved a silver-medal standard in solving problems from the International Mathematical Olympiad. This milestone demonstrates AI's growing proficiency in tackling complex mathematical challenges.

4. OpenAI's SearchGPT Challenges Google

• OpenAI has unveiled a new prototype, SearchGPT, which is positioned as a direct competitor to Google. This development signifies a major shift in the search engine landscape, with AI-driven search capabilities taking center stage.

5. Free Finetuning of GPT-4o Mini by OpenAI

• OpenAI is now offering free finetuning of its GPT-4o mini model. This move aims to make advanced AI capabilities more accessible to a broader range of users and developers.

  
  

6. Condé Nast's Legal Action Against Perplexity AI

• Condé Nast has issued a cease-and-desist order to Perplexity AI over data scraping issues. This legal action highlights the ongoing tensions between content creators and AI companies regarding data usage.

7. Increasing Web Domain Restrictions on AI Training

• A recent study has revealed a rapid increase in the number of web domains blocking AI models from accessing their data for training purposes. This trend reflects growing concerns over data privacy and the ethical use of information.

8. AI Safety Institute Legislation

• Over 45 tech organizations are advocating for the establishment of a permanent AI Safety Institute. The proposed Cantwell-Young Future of AI Innovation Act aims to codify this institute at the National Institute of Standards and Technology (NIST), ensuring the ongoing development of AI standards and guidelines.

9. AI in Energy Optimization

• AI is being increasingly used to optimize the production and distribution of renewable energy. Companies like Google’s DeepMind and IBM’s Watson are leveraging AI to enhance the efficiency and reliability of energy systems, contributing to the clean energy transition.

10. AI Ethics and Algorithmic Bias

• AI ethics remains a critical area of focus, addressing issues such as privacy, bias, transparency, and accountability. Recent discussions have highlighted the need for ethical frameworks to manage the societal impacts of AI, including the prevention of algorithmic bias and ensuring transparency in AI decision-making processes.