Microsoft Copilot Comes to Android 📱
PLUS: GPT-4 Powered Designed, GPT-4V and Gemini - The Better Vision Language Model
Today’s top AI Highlights:
Microsoft Copilot App Available on Play Store
CMU-Designed AI Coscientist That Automates Scientific Discovery
Gemini vs GPT-4V: A Comparison of Vision-Language Models
Your AI Developer that Writes Apps From Scratch While You Oversee
& so much more!
Read time: 3 mins
Latest Developments 🌍
Use GPT-4 For Free on Android 📱
Microsoft has released a dedicated Copilot app for Android on the Play Store. You wouldn’t need to have Bing separately, the Copilot app with Bing along with OpenAI’s advanced models like GPT-4 (for which you’d have to pay otherwise) and DALL.E-3.
You can ask the chatbot for research, generate images, surf the internet, create content, and use AI for your daily tasks like drafting emails, creating resumes, summarizing articles, and more!
The iOS version is not available as of now, let’s hope for hope it to come soon.
GPT-4 Carries Out Science Experiments by Itself 🧪
The first non-organic intelligent system designed to independently design, plan, and conduct complex scientific experiments! Coscientist by CMU researchers powered by GPT-4, integrates LLMs with laboratory automation technologies. It uses tools for internet and documentation search, code execution, and experimental automation. The system is designed to plan and perform complex scientific experiments autonomously.
Key Highlights:
Coscientist's architecture includes several interconnected modules, with the main module called 'Planner'. This module uses commands such as 'GOOGLE', 'PYTHON', 'DOCUMENTATION', and 'EXPERIMENT' to interact with different APIs like the Opentrons Python API and Emerald Cloud Lab's Symbolic Lab Language, and autonomously carry out intricate experiments.
It excels in tasks like planning chemical syntheses using public data, navigating through hardware documentation, executing high-level commands in cloud labs, controlling liquid handling instruments, tackling complex scientific tasks, and solving optimization problems.
In a key application, Coscientist was used to optimize Suzuki–Miyaura and Sonogashira coupling reactions. By gathering data online and writing protocols for a liquid handler, the AI system successfully performed these sophisticated chemical reactions, proving its practical value in scientific research.
Gemini vs GPT-4V: A Comparative Study 👁️🗨️
With the release of two cutting-edge multimodal models, Gemini by Google and GPT-4 Vision by OpenAI, a lot of us want to know which one of them is better at what tasks. A team of researchers put both the models to test to evaluate their linguistic as well as visual capabilities. This comprehensive evaluation covers aspects like Vision-Language Capability, Interaction with Humans, Temporal Understanding, and assessments in both Intelligence and Emotional Quotients. The overall findings reveal GPT-4V's precision and concise responses, contrasting with Gemini's detailed, expansive answers that often include imagery and links.
Key Highlights:
Basic Image Recognition: While both models excel in basic image recognition, they exhibit notable differences in text recognition and comprehension, particularly when dealing with complex formulas and table data.
Emotional Intelligence and Image Inference: Each model demonstrates the ability to understand and express emotions. However, Gemini shows a slight lag in IQ tests and object combination tasks compared to GPT-4 Vision.
Integrated Image-Text Understanding: Gemini's limitation to single image inputs sometimes places it behind GPT-4 Vision, although it holds its own in text reasoning with single images.
Industrial Application in GUI Navigation: In tasks like ergonomic keyboard selection and video content description, GPT-4 Vision generally outperforms Gemini. However, Gemini excels in providing a broader range of operational instructions and additional GUI information.
Multi-Image Recognition and Narrative Generation: GPT-4 Vision showcases superior capabilities in summarizing and identifying the content of multiple sub-images, while Gemini excels in generating coherent narratives that align with both the content and stylistic elements of the images.
Tools of the Trade ⚒️
GPT Pilot: A true AI developer who writes code, debugs it, and talks to you when it needs help. Specify the kind of app you want to build, GPT Pilot will ask questions, create product and technical requirements, set up the environment, and start coding the app step by step. It’ll ask you to review each task it finishes or to help when it gets stuck.
Pay me for my AI: Monetize your custom GPT. Start by building your own custom GPT, share it with others, and charge your users on your terms.
Inbox Zero: An open-source email app with AI automation for newsletter management, email analytics, and inbox organization. It lets you see all your subscriptions and who is sending you the most emails, the type of emails you get the most, and even automate your email replies.
Fondo TaxPlan: Makes it easy for startups to understand and manage their corporate taxes. It provides insights into how much tax a startup owes, when it's due, and where it needs to be paid. It includes a Corporate Tax Summary, a Nexus Summary, an International Activity Summary, and a Year-End Checklist.
😍 Enjoying so far, TWEET NOW to share with your friends!
Hot Takes 🔥
The newest technical papers and the oldest books are the best sources of arbitrage. They contain the least popular facts and the most monetizable truths ~ Balaji Srinivasan (Source)
The reason Sam says AGI isn't coming next year is purely financial; of course AGI is coming next year, it just won't be called ✨ AGI ✨ ~ Source
Meme of the Day 🤡
That’s all for today!
See you tomorrow with more such AI-filled content. Don’t forget to subscribe and give your feedback below 👇
Real-time AI Updates 🚨
⚡️ Follow me on Twitter @Saboo_Shubham for lightning-fast AI updates and never miss what’s trending!!
PS: I curate this AI newsletter every day for FREE, your support is what keeps me going. If you find value in what you read, share it with your friends by clicking the share button below!