• AI KATANA
  • Posts
  • Nvidia stunned the world with a ChatGPT rival that’s as good as GPT-4o

Nvidia stunned the world with a ChatGPT rival that’s as good as GPT-4o

Also: OpenAI closes $6.6 billion funding haul with investment from Microsoft and Nvidia

Author

AI KATANA
October 02, 2024

Morning!

Today’s newsletter covers a range of groundbreaking developments in AI and cloud infrastructure. Nvidia has unveiled its NVLM 1.0 models, an open-source alternative rivaling GPT-4, while OpenAI secures $6.6 billion in fresh funding, solidifying its position in the AI race. Meanwhile, Character.ai is pivoting away from AI model development following a major deal with Google, and both Oracle and Google are expanding their cloud infrastructure in Southeast Asia with multi-billion-dollar investments in Malaysia and Thailand, respectively. The tech world is clearly gearing up for the next phase of AI and cloud computing, with strategic investments reshaping the competitive landscape.

Sliced just for you:

  • 🦾 Nvidia stunned the world with a ChatGPT rival that’s as good as GPT-4o

  • 💰 OpenAI closes $6.6 billion funding haul with investment from Microsoft and Nvidia

  • 🤖 Character.ai abandons making AI models after $2.7bn Google deal

  • 🌐 Oracle to invest $6.5 billion in AI and cloud infrastructure in Malaysia

  • 🇹🇭 Google to invest $1 billion in Thai data centre, cloud infrastructure

Nvidia has launched NVLM 1.0, a family of large multimodal language models that rival GPT-4o in performance and accuracy. Rather than positioning itself as a direct competitor to ChatGPT or other AI systems like Google’s Gemini or Anthropic’s Claude, Nvidia has opted to open-source the model weights and training code, allowing developers to integrate the technology into their own AI systems. The flagship model, NVLM-D-72B, boasts 72 billion parameters and excels in both vision-language and text-only tasks, demonstrating advanced reasoning and multimodal capabilities. This move is significant because it allows smaller firms and researchers access to a high-performance AI model without the high costs associated with proprietary platforms. NVLM’s performance, particularly in multimodal understanding and text-based reasoning, positions it as a formidable open-access competitor to leading AI systems from OpenAI, Google, and Meta.

OpenAI has raised $6.6 billion in a funding round, which includes investments from key players such as Microsoft and Nvidia, potentially valuing the company at $157 billion. The capital injection, involving convertible debt from venture firms like Thrive Capital and Khosla Ventures, reflects confidence in OpenAI’s growth trajectory, particularly as it aims for $11.6 billion in revenue next year. The funding also coincides with structural changes at the company, including the departure of CTO Mira Murati. OpenAI’s rapid rise in popularity, driven by products like ChatGPT, has led to 250 million weekly active users. The company remains focused on achieving artificial general intelligence (AGI) while managing growing commercialization efforts and navigating corporate restructuring to allow for-profit conversions.

Character.ai has decided to shift its focus from developing AI models to enhancing its consumer products following a $2.7 billion deal with Google. The start-up, known for its chatbot technology that allows users to simulate conversations with various characters and celebrities, has found the cost of competing in the large language model space with tech giants like OpenAI and Google to be unsustainable. As a result, the company will prioritize refining its successful consumer offerings. The deal saw Google acquire a one-time license to Character.ai’s models and rehired its co-founders, leaving the start-up with enough funding to operate for 18 months. Despite this shift, Character.ai plans to continue AI research while exploring potential future funding and partnerships.

Oracle plans to invest over $6.5 billion to establish a public cloud region in Malaysia, aiming to support the country’s growing demand for AI and cloud services. This significant investment will enable Malaysian businesses to leverage advanced AI infrastructure and migrate critical workloads to Oracle’s cloud platform. The initiative aligns with Malaysia’s broader goals to become a digital hub, supporting the government’s plans to create 3,000 smart factories by 2030. This move expands Oracle’s presence in the Asia-Pacific region, bringing its total to 12 cloud regions, and follows similar large-scale investments by tech giants like Microsoft, Google, and Amazon in Southeast Asia. Malaysia is actively positioning itself as a leader in AI by establishing regulations and a national office to drive AI adoption across key sectors.

🇹🇭 Google to invest $1 billion in Thai data centre, cloud infrastructure

Google is set to invest $1 billion in Thailand to develop a data center and cloud infrastructure, a move aimed at meeting growing demand for cloud services and AI adoption in Southeast Asia. This investment, which will create 14,000 jobs annually through 2029, will see Google establish a data center in Chonburi and a cloud region in Bangkok. The project aligns with Thailand’s “Cloud First” policy, as the country seeks to enhance its digital infrastructure. This follows similar investments by tech giants like Microsoft, which announced its first data center in Thailand earlier this year.

🛠️ AI tools updates

Google’s new AI tool, NotebookLM, is designed to revolutionize how students, professionals, and creatives approach learning and information management. The tool functions as a personalized research assistant, capable of simplifying complex content by generating summaries, FAQs, study guides, and even audio conversations based on uploaded documents. Users can upload files or import documents from Google Drive, and NotebookLM will provide detailed insights and organizational tools, such as tables of contents and timelines, to enhance productivity and understanding. The ability to transform notes into dynamic AI-driven dialogues makes it especially useful for auditory learners. This tool aims to streamline study sessions, presentation prep, and project management, providing a more interactive and engaging way to process information.

💵 Venture Capital updates

AI chipmaker Cerebras Systems has filed for an IPO, planning to trade under the ticker symbol “CBRS” on the Nasdaq. The company, which competes with Nvidia, reported a net loss of $66.6 million on $136.4 million in sales for the first half of 2024. Known for its WSE-3 chip, which boasts more cores and memory than Nvidia’s H100, Cerebras also offers cloud services using its proprietary computing clusters. Despite growing competition from giants like Microsoft, Google, and Amazon, Cerebras remains focused on AI infrastructure, with Group 42, a UAE-based AI firm, accounting for 83% of its revenue in 2023. The IPO will be led by Citigroup and Barclays, with key investors including Foundation Capital, Benchmark, and Eclipse Ventures.

🫡 Meme of the day

⭐️ Generative AI image of the day