🍓 Rumors of OpenAI's Newest Model

Author

AI Breakfast
August 09, 2024

Good morning. It’s Friday, August 9th.

Did you know: On this day in 1991, the first email from space was sent via a Macintosh Portable.

In today’s email:

  • OpenAI’s

  • Foundation Model

  • ChatGPT's "Secret Ingredient" May Be Holding Back LLMs

  • Prompt Poet

  • Apple Intelligence Update

  • 5 New AI Tools

  • Latest AI Research Papers

You read. We listen. Let us know what you think by replying to this email.

Today’s trending AI news stories

Sam Altman Stokes Rumors of New OpenAI Foundation Model 'Strawberry'

Sam Altman's recent post on X, featuring a picture of strawberries with the caption “I love summer in the garden,” has sparked a frenzy of speculation about OpenAI's rumored new foundation model, code-named “Strawberry.” This seemingly innocuous post has excited the AI community, suggesting potential confirmation of OpenAI's next major development, possibly the anticipated GPT-5. This follows whispers of a more ambitious project, codenamed "Q*," aiming for AGI—a digital system capable of outperforming humans in economically significant tasks.

Adding to the intrigue is a new anonymous chatbot model appearing on the open-source LMSYS Arena, reportedly exhibiting advanced reasoning beyond GPT-4o. This new development follows nearly 17 months since OpenAI released GPT-4, with competitors like Anthropic, Google, and Meta gaining ground. The AI community watches with bated breath as the company attempts to maintain its lead in the AI race. Read more.

AI Researcher Says ChatGPT's "Secret Ingredient" May Be Holding Back LLM Capabilities

Andrej Karpathy, a distinguished AI researcher and former OpenAI scientist, has scrutinized the current method of training language models, Reinforcement Learning from Human Feedback (RLHF). Karpathy likens this technique to teaching a parrot to parrot human speech, critiquing it for relying on human input to direct AI behaviour instead of nurturing true intelligence.

He draws a stark contrast between RLHF and the "real RL" used to train DeepMind's AlphaGo, which involved the AI learning through self-play, optimizing its performance by simulating thousands of games, and developing superhuman capabilities without direct human intervention.

Karpathy contends that RLHF is essentially a "vibe check," aligning AI behavior with human preferences rather than equipping it with the tools for independent thought. He warns that this reliance on human guidance may limit the AI's potential to evolve into a truly autonomous problem solver. Read more.

Meet Prompt Poet: The Google-acquired tool revolutionizing LLM prompt engineering

Prompt Poet, acquired by Google from Character.ai, is redefining prompt engineering for large language models (LLMs). This tool simplifies the creation of sophisticated prompts through a low-code interface, making it accessible to both technical and non-technical users. It integrates YAML and Jinja2 for flexible prompt templates and efficient context management, allowing users to incorporate external data seamlessly. In practice, Prompt Poet streamlines prompt design by managing complex instructions and dynamic data, reducing manual input errors and enhancing model responses. Tested with various scenarios, such as daily planning assistants, Prompt Poet demonstrates significant improvements in context-aware output. This development not only showcases a leap forward in AI prompt engineering but also hints at Google’s shift towards more intuitive and accessible AI tools. Read more.

Apple Could Charge Up to $20 for Some Apple Intelligence AI Features, Analysts Say

Apple is poised to introduce its generative AI model, Apple Intelligence, with analysts predicting a premium tier carrying a $20 price tag. Slated for a late-year launch, Apple Intelligence aims to augment Siri's capabilities and introduce generative features like automatic email and image creation.

While the rollout will initially exclude some regions, a subscription model may be employed for premium features. Analysts suggest that Apple might incorporate these features into the existing Apple One subscription, which costs $19.95 per month and bundles various services. Apple’s services division, which generated $24.2 billion in the last quarter, could leverage this new AI system to further monetize its offerings. The strategy aligns with industry trends, where companies like OpenAI and Microsoft also charge for advanced AI services. Apple’s move is seen as a method to deepen user engagement and generate additional revenue. Read more.

Etcetera: Stories you may have missed

5 new AI-powered tools from around the web

AICamp offers businesses a secure platform to access, customize, and deploy AI models, including GPT-4o, Claude, and Gemini enhancing collaboration and data control.

Inkeep converts documentation into AI-powered tools like search bars, chatbots, and Slack integrations, facilitating precise user support and information retrieval.

Omnifact provides ultra-secure AI chat assistants, supporting custom deployments and internal knowledge integration for enterprise teams, prioritizing data privacy control.

PolygrAI uses AI to perform real-time deception risk assessment and sentiment analysis by evaluating behavioral cues during meetings and interviews.

Text2Infographic converts text into visually engaging infographics, conducting research, sourcing images, and citing references to enhance content presentation.

arXiv is a free online library where researchers share pre-publication papers.

Thank you for reading today’s edition.

Your feedback is valuable. Respond to this email and tell us how you think we could add more value to this newsletter.

Interested in reaching smart readers like you? To become an AI Breakfast sponsor, reply to this email!