Google DeepMind’s LLM Solves Complex Math
Plus: OpenAI's Prompt Engineering Guide, ByteDance secretly uses OpenAI’s Tech.
Hello Engineering Leaders and AI Enthusiasts!
Welcome to the 170th edition of The AI Edge newsletter. This edition brings you: Google DeepMind’s new LLM ‘FunSearch’ which solves complex math problems.
And a huge shoutout to our incredible readers. We appreciate you😊
In today’s edition:
🧮 Google DeepMind’s LLM solves complex math
📘OpenAI released its Prompt Engineering Guide
🤫 ByteDance secretly uses OpenAI’s Tech
📚 Knowledge Nugget: Role-playing with AI will be a powerful tool for writers and educators by
Let’s go!
Google DeepMind’s LLM solves complex math
Google DeepMind has used an LLM called FunSearch to solve an unsolved math problem. FunSearch combines a language model called Codey with other systems to suggest code that will solve the problem. After several iterations, FunSearch produced a correct and previously unknown solution to the cap set problem.
This approach differs from DeepMind's previous tools, which treated math problems as puzzles in games like Go or Chess. FunSearch has the advantage of finding solutions to a wide range of problems by producing code, and it has shown promising results in solving the bin packing problem.
Why does this matter?
FunSearch's ability to solve an unsolved math problem showcases AI matches high-level human skills in several ways. Its advances in core reasoning abilities for AI, such as displayed by FunSearch, will likely unlock further progress in developing even more capable AI. Together, these interrelated impacts mean automated math discoveries like this matter greatly for advancing AI toward more complex human thinking.
OpenAI released its Prompt Engineering Guide
OpenAI released its own Prompt Engineering Guide. This guide shares strategies and tactics for improving results from LLMs like GPT-4. The methods described in the guide can sometimes be combined for greater effect. They encourage experimentation to find the methods that work best for you.
The OpenAI Platform provides six strategies for getting better results with language models. These strategies include writing clear instructions, providing reference text, splitting complex tasks into simpler subtasks, giving the model time to think, using external tools to compensate for weaknesses, and testing changes systematically. By following these strategies, users can improve the performance and reliability of the language models.
Why does this matter?
Releasing an open prompt engineering guide aligns with OpenAI's mission to benefit humanity. By empowering more people with skills to wield state-of-the-art models properly, outcomes can be directed toward more constructive goals rather than misuse - furthering responsible AI development.
ByteDance secretly uses OpenAI’s Tech
ByteDance, the parent company of TikTok, has been secretly using OpenAI's technology to develop its own LLM called Project Seed. This goes against OpenAI's terms of service, prohibiting the use of their model output to develop competing AI models.
Internal documents confirm that ByteDance has relied on the OpenAI API for training and evaluating Project Seed. This practice is considered a faux pas in the AI world, and Microsoft, through which ByteDance accesses OpenAI, has the same policy
Why does this matter?
ByteDance's use of OpenAI's tech highlights the intense competition in the generative AI race. Ultimately, this case highlights the priority of integrity and transparency in progressing AI safely.
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: Role-playing with AI will be a powerful tool for writers and educators
The author
discusses the potential of using generative AI as a tool for writers and educators. They have tested AI simulations on the content of their books and found it to be a useful tool for exploring alternative versions of the world and enhancing understanding of historical periods.
(An anonymous survey about how AI activities impacted their learning)
While accuracy is challenging, the author believes that generative AI can help kickstart historical imagination and empathy. They provide examples of simulated experiences and visuals that have been helpful in their own writing process.
Why does this matter?
Using AI to simulate historical contexts and perspectives, writers and educators can spark imagination that deepens insights into the past. While acknowledging the need to address inaccuracies, the author shows the promise of AI as a supplementary tool to expand human knowledge and empathy.
What Else Is Happening❗
💡 Deloitte is turning towards AI to avoid mass layoffs in the future
The company plans to use AI to assess the skills of its existing employees and identify areas where they can be shifted to meet demand. This move comes after Deloitte hired 130,000 new staff members this year but warned thousands of US and UK employees that their jobs were at risk of redundancy due to restructuring. (Link)
🌐 Ola’s founder have announced an Indian LLM
This new multilingual LLM will have generative support for 10 Indian languages and will be able to take inputs in a total of 22 languages. It has been trained on over two trillion tokens of data for Indian languages. And will be trained on 'Indian ethos and culture'. The company will also develop data centers, supercomputers for AI, and much more. (Link)
🧸 Grimes partnered with Curio Toys to create AI toys for children
Musician Grimes has partnered with toy company Curio to create a line of interactive AI plush toys for children. The toys, named Gabbo, Grem, and Grok, can converse with and "learn" the personalities of their owners. The toys require a Wi-Fi connection and come with an app that provides parents with a written transcript of conversations. (Link)
🔧 Agility uses LLMs to enhance communication with its humanoid robot- Digit
The company has created a demo space where Digit is given natural language commands of varying complexity to see if it can execute them. The robot is able to pick up a box of a specific color and move it to a designated tower, showcasing the potential of natural language communication in robotics. (Link)
🍔 CaliExpress is hailed as the world’s first autonomous AI restaurant
The eatery, set to open before the end of the year, will feature robots that can make hamburgers and French fries. However, the restaurant will still have human employees who will pack the food and interact with customers. (Link)
That's all for now!
If you are new to The AI Edge newsletter, subscribe to get daily AI updates and news directly sent to your inbox for free!
Thanks for reading, and see you tomorrow. 😊