Weekly Tech News of July Last Week 2024

Gen-3 Alpha—The New Generation of Runway AI

The new update called “Gen-3 Alpha” from “Runway” offers advanced Text to Video generating capabilities. Where, you are enable to create videos by just entering the text prompts and adjusting settings. 

Gen-3 Alpha introduce the improved fidelity, consistency, and motion over it’s last launched called Gen-2. The steps include in Gen-3 Alpha on entering a text prompt result in generating videos up to 10 seconds and iterating for diverse results. 

This upgradation of Runway supports various tools like Motion Brush and Director Mode to provide a fine-grained control over video elements. 

Utilizing the excels of Gen-3 Alpha, you can generate photorealistic humans and can customize the consistent characters and artistic requirements. This tool includes new safeguards which was developed collaboratively by scientists, engineers, and artists.

Official Source: https://runwayml.com/ai-tools/gen-3-alpha/

What New Meta Has Brought to You: Explore Llama 3.1, The Next Gen-AI

Llama 3.1, an open-source AI model by Meta has been recently launched and has become the talk of the town in AI universe. The only one that isd available in 8B, 70B, and 405B versions and designed for fine-tuning, distilling, and deploying. 

This 405B model supports diverse use cases and advanced applications by having the key features include real-time and batch inference, fine-tuning, model evaluation, safety guardrails, and synthetic data generation. 

Llama 3.1 is excel in performing multilingual tasks including the tool use, reasoning, code benchmarks, and so on. In addition to this, it supports ten languages and includes resources for advanced use cases. 

It’s Llama Guard 3 ensures safety  of the model and take care that the model is accessible on platforms like GitHub, Hugging Face, and Kaggle. Plus, it’s demo videos and comprehensive documentation are also available on the official site for a better execution guide.

Official Source: https://llama.meta.com/

OpenAI‘s GPT-4o Mini Model — An Innovative Upsurge in AI (Cost Efficient)

OpenAI has launched GPT-4o mini, a highly cost-effective small AI model priced at 15 cents per million input tokens and 60 cents per million output tokens. It surpasses GPT-3.5 Turbo in benchmarks along with the chat preferences. 

GPT-4o mini Supports text, vision, and future multimodal inputs and also excels in reasoning, math, and executing the coding tasks very smoothly. This upgrade of OpenAI offers a 128K token context window for real-time text responses that assist in enhancing the customer support and data analysing process. 

With built-in safety measures of GPT-4o mini, you are allos to use it’s APIs feature. Apart from this, if you are a ChatGPT user then you can access it even by today. As OpenAI aims to make AI more accessible, reliable, and integrated into daily digital experience hance released GPT-4o mini.

Official Source: https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/

OpenAI’s SearchGPT: A Prototype of Searching Universe

Meet SearchGPT which is recently testing by OpenAI, a prototype that ais in enhancing AI’s  search within real-time web information and offers fast answers with clear sources. 

SearchGPT is not available publicly yet as it has been given to a small group of user for feedback by having the aim of integrating with the best features into ChatGPT. To provide a better transparency, SearchGPT promises to deliver timely, relevant answers, citing linking sources, and the list goes on. 

This new launch is also partnering with publishers to enhance the user discovery while respecting the journalism. The prototype of SearchGPT allows publishers to manage their appearance in searches. Moreover, Feedback is welcomed as OpenAI continues to improve the service and Interested users can join the waitlist to get the early access.

Official Source: https://openai.com/index/searchgpt-prototype/

Related Post

Weekly Tech News of July second Week 2024