OpenAI's Secret AI: Project Strawberry
Plus: Meta researchers developed "System 2 distillation" for LLMs, Amazon's Rufus AI is now available in the US.
Hello Engineering Leaders and AI Enthusiasts!
Welcome to the 317th edition of The AI Edge newsletter. This edition features OpenAI’s new secret AI tech named Project Strawberry.
And a huge shoutout to our amazing readers. We appreciate you😊
In today’s edition:
🍓 OpenAI is working on an AI codenamed "Strawberry"
🧠 Meta researchers developed "System 2 distillation" for LLMs
🛒 Amazon's Rufus AI is now available in the US
📚 Knowledge Nugget: In AI we trust, part II by
Let’s go!
OpenAI is working on an AI codenamed "Strawberry"
The project aims to improve AI's reasoning capabilities. It could enable AI to navigate the internet on its own, conduct "deep research," and even tackle complex, long-term tasks that require planning ahead.
The key innovation is a specialized post-training process for AI models. The company is creating, training, and evaluating models on a "deep-research" dataset. The details about how previously known as Project Q, Strawberry works are tightly guarded, even within OpenAI.
The company plans to test Strawberry's capabilities in conducting research by having it browse the web autonomously and perform tasks normally performed by software and machine learning engineers.
Why does it matter?
If successful, Strawberry could lead to AI that doesn't just process information but truly understands and reasons like humans do. And may unlock abilities like making scientific discoveries and building complex software applications.
Meta researchers developed "System 2 distillation" for LLMs
Meta researchers have developed a "System 2 distillation" technique that teaches LLMs to tackle complex reasoning tasks without intermediate steps. This breakthrough could make AI applications zippier and less resource-hungry.
This new method, inspired by how humans transition from deliberate to intuitive thinking, showed impressive results in various reasoning tasks. However, some tasks, like complex math reasoning, could not be successfully distilled, suggesting some tasks may always require deliberate reasoning.
Why does it matter?
Distillation could be a powerful optimization tool for mature LLM pipelines performing specific tasks. It will allow AI systems to focus more on tasks they cannot yet do well, similar to human cognitive development.
Amazon's Rufus AI is now available in the US
Amazon's AI shopping assistant, Rufus is now available to all U.S. customers in the Amazon Shopping app.
Key capabilities of Rufus include:
Answers specific product questions based on product details, customer reviews, and community Q&As
Provides product recommendations based on customer needs and preferences
Compares different product options
Keeps customers updated on the latest product trends
Accesses current and past order information
This AI assistant can also tackle broader queries like "What do I need for a summer party?" or "How do I make a soufflé?" – proving it's not just a product finder but a full-fledged shopping companion.
Amazon acknowledges that generative AI and Rufus are still in their early stages, and they plan to continue improving the assistant based on customer feedback and usage.
Why does it matter?
Rufus will change how we shop online. Its instant, tailored assistance will boost customer satisfaction and sales while giving Amazon valuable consumer behavior and preferences insights.
Enjoying the daily updates?
Refer your pals to subscribe to our daily newsletter and get exclusive access to 400+ game-changing AI tools.
When you use the referral link above or the “Share” button on any post, you'll get the credit for any new subscribers. All you need to do is send the link via text or email or share it on social media with friends.
Knowledge Nugget: In AI we trust, part II
Legal expert
found that the Claude 3 Opus correctly decided 27 out of 37 Supreme Court cases from this term. Claude AI showed an uncanny ability to create novel legal standards and analyze complex issues, often matching or surpassing the insights of human Supreme Court clerks.Moreover, Claude showed proficiency in spotting methodological errors in expert testimony and proposing creative solutions to legal challenges.
Why does it matter?
Claude accomplishes these tasks about 5,000 times faster than human law clerks; this shows that AI could significantly improve the efficiency and accuracy of legal research and preliminary case evaluations.
What Else Is Happening❗
🤖 OpenAI rushed safety tests for GPT-4 Omni
OpenAI is under scrutiny for allegedly rushing safety tests on its latest model, GPT-4 Omni. Despite promises to the White House to rigorously evaluate new tech, some employees claim the company compressed crucial safety assessments into a week to meet launch deadlines. (Link)
📣 OpenAI whistleblowers filed a complaint with the SEC
They allege the company's NDAs unfairly restrict employees from reporting concerns to regulators. This complaint, backed by Senator Chuck Grassley, calls for investigating OpenAI's practices and potential fines. (Link)
🧠 DeepMind introduces PEER for scaling language models
Google DeepMind introduced a new technique, "PEER (Parameter Efficient Expert Retrieval)," that scales language models using millions of tiny "expert" modules. This approach outperforms traditional methods, achieving better results with less computational power. (Link)
✍️Microsoft is adding handwriting recognition to Copilot in OneNote
The feature can read, analyze, and convert handwritten notes to text. Early tests show impressive accuracy in deciphering and converting handwritten notes. It can summarize notes, generate to-do lists, and answer questions about the content. It will be available to Copilot for Microsoft 365 and Copilot Pro subscribers. (Link)
🆕Rabbit R1 AI assistant adds a Factory Reset option to wipe user data
Rabbit's R1 AI assistant was storing users' chat logs with no way to delete them. But a new update lets you wipe your R1 clean. The company also patched a potential security hole that could've let stolen devices access your data. (Link)
New to the newsletter?
The AI Edge keeps engineering leaders & AI enthusiasts like you on the cutting edge of AI. From machine learning to ChatGPT to generative AI and large language models, we break down the latest AI developments and how you can apply them in your work.
Thanks for reading, and see you tomorrow. 😊