Live
- PM Modi highlights govt's efforts to make Odisha prosperous and one of the fastest-growing states
- Hezbollah fires 200 rockets at northern, central Israel, injuring eight
- Allu Arjun's Family Appearance on Unstoppable with NBK Breaks Viewership Records
- Unity of hearts & minds essential for peace & progress, says J&K Lt Governor
- IPL 2025 Auction: I deserve Rs 18 cr price, says Chahal on being acquired by Punjab Kings
- EAM Jaishankar inaugurates new premises of Indian embassy in Rome
- Sailing vessel INSV Tarini embarks on second leg of expedition to New Zealand
- Over 15,000 people affected by rain-related disasters in Sri Lanka
- IPL 2025 Auction: RCB acquire Hazlewood for Rs 12.50 cr; Gujarat Titans bag Prasidh Krishna at Rs 9.5 crore
- Maharashtra result reflects the outcome of Congress' destructive politics: BJP's Shazia Ilmi
Meta to build an advanced AI Model to Rival OpenAI's GPT-4
Mark Zuckerberg's Meta is building a new AI model that will be more powerful than its current Llama 2 model and rival the capabilities of OpenAI's GPT-4.
In July, Meta, the parent company of Facebook, Instagram and WhatsApp, publicly released its AI model Llama 2. Built on 65 billion parameters, Meta described its platform as a "smaller fundamental model" suitable for AI research and testing new approaches as it consumed less computing power. But now, a new report suggests that CEO Mark Zuckerberg is entering the big AI race and intends to compete with companies like OpenAI. Meta is working on a new AI model far superior to Llama 2 and will rival GPT-4, OpenAI's latest large language model.
As per a Wall Street Journal report, Meta wants its new AI model ready for launch next year and hopes to start training the AI model as early as the first three months of 2024. Interestingly, Zuckerberg is pushing to make the AI Model open source and therefore free for companies to create AI tools, just like its predecessor.
The new AI model is being built to have all the capabilities of an LLM, including sophisticated text generation, analysis and human expression. According to the report, Meta has started purchasing more Nvidia H100 AI training chips, the most advanced chip produced by the company. Meta wants to train the new model on its infrastructure, unlike Llama 2, which was introduced on Microsoft's Azure cloud computing platform.
Meta to rival OpenAI with its GPT 4-like AI model
Previously, it was reported that Meta created a team of AI researchers to create tools that can provide a human-like experience. Now, Zuckerberg is said to have accelerated the team's work. According to some leaks, this team has already started creating artificial intelligence tools for the company. In June, a report from The Verge claimed that Instagram was working on a chatbot with 30 different human personalities. This chatbot could launch later this month as Personas, a chatbot the company has previously promoted. Read more: ChatGPT Enterprise found, ready to compete with Microsoft Bing Enterprise.
Zuckerberg will also attend AI-related conclaves to discuss how AI can be handled. He will attend the much-discussed summit hosted by US Senator Chuck Schumer and OpenAI's Sam Altman, Google CEO Sundar Pichai, and other top tech executives.
