AI News: How Claude 3 Opus Beat ChatGPT’s GPT-4 on Chatbot Arena

TL;DR

Claude 3 Opus overtook GPT-4 as Chatbot Arena’s top-performing language model based on user votes. The week’s other AI news included Amazon adding $2.75 billion to its Anthropic investment, Databricks releasing the open-source DBRX model, Stability AI’s CEO stepping down, and OpenAI beginning a pilot program for GPT earnings based on usage. Read on for specific comparisons, costs, and access details.

Transcript

there's a lot of really fun and fascinating stuff happening in the world of AI this week and I don't want to waste any time so let's get right into it starting with the fact that we have a new toplarge language model I've talked about Claude 3 Opus in past videos well it is officially the Top Model according to most benchmarks chatbot arena is this... Read More

Key Insights

• ❓ Cloze 3 Opus surpasses GPT-4 as the top language model, highlighting the continuous progress in natural language processing.
• ❓ Amazon's increased investment in anthropic reflects the company's ambition to become a major player in the AI market.
• 🤗 Data Bricks' new language model, dbrx, showcases the ongoing advancements in open-source AI models.
• 💱 Stability AI's leadership changes and release of Stable Code Instruct 3B indicate a challenging period for the company.
• 💗 OpenAI's efforts to monetize the GPT store signal the growing importance of AI-generated content.
• 🎨 Adobe Gen Studio aims to provide a centralized platform for designing and managing creative assets.
• 🔨 New AI tools like Hume and Bezi AI offer unique features such as emotional intelligence and text-to-3D object generation.

Questions & Answers

Q: Which chatbot finally beat ChatGPT’s GPT-4 on Chatbot Arena?

Claude 3 Opus surpassed GPT-4 to become Chatbot Arena’s top-performing model. Chatbot Arena is a crowdsourced platform where people use different chatbots and vote for their favorites.

Q: How much does Claude 3 Opus cost, and what can’t it do compared with ChatGPT?

Claude 3 Opus is Claude’s premium model and costs $20 per month. Unlike ChatGPT, Claude does not generate images or search the web, although the speaker considers it better for nearly every other use case.

Q: How much has Amazon invested in Anthropic?

Amazon initially invested $1.25 billion in Anthropic and announced another $2.75 billion investment. Amazon will retain a minority stake and will not receive a seat on Anthropic’s board.

Q: How does Databricks’ DBRX compare with other open-source language models?

According to benchmark tests cited in the transcript, DBRX beat Llama 2, Mixtral, and Grok 1 in language understanding, programming, and math. It also outperformed the free version of GPT-3.5 on every benchmark shown.

Q: How can users download the DBRX model weights?

The DBRX model weights are available through Hugging Face. Users must submit an application and receive approval before they can download them.

Q: Why did Stability AI CEO Emad Mostaque step down?

The exact reason remained unclear, although there was speculation that he had been forced out amid employee departures and rumors of financial trouble or possible bankruptcy. Mostaque said he stepped down because concentrating power in AI was harmful and that he was going “all in” on decentralized AI.

Q: What did Stability AI release after its CEO stepped down?

Stability AI announced Stable Code Instruct 3B the day after Emad Mostaque stepped down. It is a new coding-focused large language model, and its release suggested that the company intended to continue shipping products despite the leadership change.

Q: What other AI product and monetization updates were announced?

Elon Musk said all Premium subscribers on X would gain access to Grok, which had previously been available through a $20-per-month Premium Plus subscription. OpenAI also began a pilot program that would pay GPT builders based on usage.

Summary & Key Takeaways

• Cloze 3 Opus, a language model, has been voted the top model on Chatbot Arena, surpassing GPT-4 in performance.

• Amazon announced an additional investment of $2.75 billion in anthropic, following a similar deal with OpenAI.

• Data Bricks released a new language model called dbrx, outperforming other open-source models in various domains.

• Stability AI's CEO stepped down amid rumors of financial issues, while the company introduced a new language model, Stable Code Instruct 3B.