Meta releases the biggest and best open-source AI model yet

Pivot 5: 5 stories. 5 minutes a day. 5 days a week.

1. Meta releases the biggest and best open-source AI model yet

Meta is releasing Llama 3.1, the largest-ever open-source AI model, which claims outperforms GPT-4o and Anthropic's Claude 3.5 Sonnet on several benchmarks. The company is making the Llama-based Meta AI assistant available in more countries and languages and adding a feature that can generate images based on someone's specific likeness.

CEO Mark Zuckerberg predicts that Meta AI will be the most widely used assistant by the end of this year, surpassing ChatGPT. Meta is working with over two dozen companies to help developers deploy their own versions, with Llama 3.1 costing roughly half that of OpenAI's GPT-4o.

Read the full story here

2. Google creates AI weather simulator that measures climate change

Google has developed an AI-enhanced weather simulator, the NeuralGCM model, which accurately predicts warming from climate change. The hybrid model combines traditional forecasts with machine learning, achieving comparable spatial bias to global cloud-resolving models.

The model could predict hurricanes in the future, and researchers are working on a year-ahead hurricane prediction feature.

Read the full story here

3. Nebius, a ‘startup’ with plans to be a European AI compute leader

Nebius, a fledgling AI infrastructure business, has emerged from the ashes of Yandex, a multibillion-dollar juggernaut once touted as the "Google of Russia." Arkady Volozh, former CEO of Yandex, said the company is similar to a startup but an "unusually big one."

Volozh was forced out of Yandex in 2022 after the European Union placed him on a sanctions list. The EU removed Volozh from the list in March this year, paving the way for his return as CEO of Yandex's next incarnation, one whose team and data centers are entirely outside Russia.

Read the full story here

4. Apple shows off open AI prowess: new models outperform Mistral and Hugging Face offerings

Apple has released a family of open DCLM models on Hugging Face, including two models with 7 billion and 1.4 billion parameters. The larger model outperforms Mistral-7B and is close to other leading open models like Llama 3 and Gemma.

The project, led by a team of multidisciplinary researchers, focuses on designing high-quality datasets for training AI models, particularly in the multimodal domain. The 7B model, trained on 2.5 trillion tokens, delivers 63.7% 5-shot accuracy on MMLU, a 6.6 percentage point improvement compared to MAP-Neo.

Read the full story here

5. AI Startup Cohere Valued at $5.5 Billion in New Funding Round

Cohere, a Canadian AI startup, has raised $5.5 billion in a $500 million round, raising its total funding to $970 million. The company focuses on practical goals, creating large language models from scratch, rather than artificial general intelligence (AGI).

Cohere has clients in various industries, including retailers, tech companies, and banks. However, research by PYMNTS Intelligence shows a gap between perceived potential and actual application.

Read the full story here

Advertise with Pivot 5 to reach influential minds & elevate your brand

Get your brand in front of 65,000+ businesses and professionals who rely on Pivot 5 for daily AI updates. Book future ad spots here.