Microsoft GPT Rival + OpenAI Lobbied the EU

Meet Orca, Microsoft’s new 13 billion parameter AI model that can imitate GPT-4

Microsoft Research just introduced a new AI model called Orca, which learns by imitating large language models like GPT-4.

Orca addresses the limitations of smaller models by emulating the reasoning processes of larger foundation models. This approach allows Orca to be optimized for specific tasks and trained using GPT-4 while requiring fewer computing resources to operate. It offers researchers the flexibility to optimize their models according to their needs and run them independently without relying on a large data center.

Orca, with 13 billion parameters, can imitate and learn from GPT-4. Microsoft utilizes extensive imitation data to enable progressive learning with Orca, It also outperforms conventional AI models by 42% on AGIEval.

Despite its smaller size, Orca performs comparably to ChatGPT in reasoning benchmarks like BBH and demonstrates competitive performance on academic exams such as SAT, LSAT, GRE, and GMAT, although it falls behind GPT-4.

The Microsoft research team expects Orca to further enhance its skills and capabilities by learning from step-by-step explanations created by humans and more advanced language models.

Read More

Research Paper

OpenAI Lobbied the E.U. to Water Down AI Regulation

OpenAI, in conjunction with CEO Sam Altman's global tour advocating for AI regulation, has been revealed to have lobbied for the reduction of regulatory burdens on the company within the European Union's (EU) proposed AI Act. Documents obtained from the European Commission indicate that OpenAI proposed amendments to the AI Act that would exclude its general-purpose AI systems, such as GPT-3 and Dall-E 2, from being classified as "high risk" and subject to stricter regulations.

Microsoft and Google have also supported easing the regulatory requirements on large AI providers. The lobbying efforts seem to have been successful, as the final draft of the AI Act did not categorize general-purpose AI systems as inherently high risk. OpenAI's engagement with EU officials shows its desire to mitigate overregulation and maintain a flexible regulatory framework that aligns with its financial interests.

However, concerns have been raised about the potential impact on AI innovation and the need for effective regulation. OpenAI continues to engage with policymakers and expresses its commitment to ensuring the safe development and use of AI technologies.

Read More

In Other News

China's ByteDance Has Gobbled Up $1 Billion of Nvidia GPUs for AI This Year

Best-selling newspaper in Germany to replace editors with AI

How AI is helping us talk to animals

10 ChatGPT Prompts to Make You 10x Smarter

Prompt:

Copy/paste the prompts here: https://www.promptvibes.com/view-and-test?recordId=reckrzAuAIXTi5WrG

Prompt for formulating better thoughts

Prompt:

Copy/paste the prompts here: https://www.promptvibes.com/view-and-test?recordId=recN7zRen5L71g5sz

Chef's Request

Fictional Recipe Generator w/ Image Generator

Prompt:

Copy/paste the prompts here: https://www.promptvibes.com/view-and-test?recordId=rec5dg7unhcLBqLpG

Do you wonder if your AI-generated text is correct? Paste it in the Factiverse editor, and leave it up to our AI to find factual mistakes. We will also search Google and Bing in real-time to provide links to credible sources and save you hours of research.

In a secret forest testing ground, squirrel military geniuses assemble a tank made entirely out of hazelnuts