Apple Launches AI-Powered M4 Chip

PLUS: GPT-2 chatbot is back again, OpenAI's damage control measures

Today’s top AI Highlights:

  1. Apple gives AI the center stage for the first time at its iPad event

  2. GPT2 chatbot appears again on the Lmsys Chatbot Arena with two hilarious names

  3. OpenAI does some disaster management after getting sued for copyright infringement

  4. 100% free AI workflows to automate any task

  5. Build awesome LLM apps with RAG using OpenAI, Anthropic, Gemini and opensource models

& so much more!

Read time: 3 mins

Latest Developments 🌍

The New M4 Chip for Powering Next-gen AI Apps 🚀

Apple finally wrapped up its special event to release the new iPad line-up. This was probably the first time the company boasted the “AI” word and emphasized how they have integrated AI across various features. They also released the much-anticipated M4 chip with fundamental improvements to the CPU, GPU, Neural Engine, and memory system to power the next-gen AI apps on Apple’s devices.

Key Highlights:

  1. M2 chip in iPad Air: The new iPad Air now boasts the M2 chip, previously exclusive to iPad Pro, with faster CPU and GPU. This significantly boosts ML tasks like identifying and isolating objects and texts and images faster and smoother.

  2. Third-party Apps: While Apple hasn’t trained any AI model in-house to showcase, it mentioned Photomator that uses AI models trained on millions of images to enhance photos with a single click.

  3. M4 Chip for Next-gen AI: The new iPad Pro has the groundbreaking M4 chip with an even more powerful Neural Engine capable of 38 trillion operations per second. This allows for complex AI applications like isolating subjects from video backgrounds with incredible speed.

  1. Adaptive Flash: Scanning documents becomes easier with the new AI-powered flash in iPad Pro. It automatically detects documents and adjusts flash settings to minimize shadows, resulting in significantly clearer scans.

  2. AI for Video Production: Final Cut Pro on iPad receives a major update with AI-powered features like Live Multicam, to connect and synchronize up to four cameras simultaneously, transforming your iPad into a multicam production studio.

  3. AI for Music Production: The Logic Pro app gets new AI features like AI-powered keyboard player and drummer for backing your music, adding analog warmth to digital instruments, and isolating vocals and different instruments from any recording for remixing.

GPT2 Chatbot Appears Again! 🤖

Last week a new chatbot called the “gpt2-chatbot” appeared on LMSys’ chatbot arena and AI enthusiasts went boggle on it. With its impressive reasoning capabilities, it was speculated that gpt2-chatbot is by OpenAI and it is testing its new AI chatbot in stealth. It was unavailable the very next day.

Yesterday, two new gpt2 chatbots appeared on the Arena - “im-a-good-gpt2-chatbot” and “im-also-a-good-gpt2-chatbot.” Testing them on every kind of possible task, both chatbots gave excellent responses with comprehensive and structured reasoning. For instance,

  1. Both the chatbots created the Flappy Bird game’s code perfectly in one shot.

  2. A Reddit user gave im-also-a-good-gpt2-chatbot a freshman physics problem that it solved perfectly while no other LLM was able to do it.

  3. Another X user gave the chatbot a rubric inquiry where many LLMs stumble but im-also-a-good-gpt2-chatbot gave the right answer with a perfect reasoning structure.

  4. Check this compilation for more such examples

It’s not easy to chat with these two chatbots. Right now, they are only available in the “Battle Mode” of the Arena. Here’s how you might get lucky:

  1. Visit the Lmsys chatbot arena 

  2. Switch to Arena (battle) mode

  3. Type your prompt

  4. Rate the responses. Check the bottom of the chatbox to see which model you interacted with. You will find either of the two chatbots.

  5. If you don’t find the gpt2-chatbot, refresh and try again!

Who has created these chatbots? Take it with a grain of salt but it seems these are OpenAI’s. First, Sam Altman tweeted its name before these chatbots appeared. Second, check this error in the below screenshot.

OpenAI for Transparency and Accountability in AI Media 🧐

Generative AI tools like DALL·E 3 and ChatGPT have opened doors to incredible creativity and efficiency, but have also risen concerns about the authenticity of content. This becomes especially crucial in sensitive times like elections. So understanding the source of AI-generated images, videos, and audio becoes crucial. OpenAI is taking steps to address these concerns and promote transparency through collaborations and their own proprietary technology.

Key Highlights:

  1. C2PA: OpenAI is joining Coalition for Content Provenance and Authenticity’s (C2PA) committee. C2PA is a widely used standard for digital content certification to prove the source of the content.

  2. C2PA Integration: DALL·E 3 images already include C2PA metadata, and the upcoming text-to-video model, Sora, will also incorporate this, making it difficult to fake or alter information about content origin.

UI interface
  1. Societal Resilience Fund: OpenAI and Microsoft are partnering to launch a $2 million fund to support AI education and understanding.

  2. Provenance Methods: OpenAI is developing tamper-resistant watermarking for audio and image content, and a classifier to distinguish content generated by DALL.E 3. This tool can shows an accuracy of ~98%. Applications for getting access to this classifier are open till July 31.

Besides content authentication, OpenAI is also taking steps to distance itself from copyright infringement allegations for training their AI models. In August 2023, they had given an option to content creators to turn off OpenAI’s bot from scraping the data from their websites. They have also partnered with companies like Stack Overflow, Associated Press, Financial Times, and Le Monde to access their data for training their AI models.

OpenAI is building a one-of-its-kind Media Manager tool through which creators and content owners can tell OpenAI what they own, and specify how they want their works to be included or excluded from AI training. Slated to release by 2025, this would give more flexibility to content owners than the GPTBot, and might also enable the AI models to clearly state the content/copyright owner in the output.

😍 Enjoying so far, share it with your friends!

Tools of the Trade ⚒️

  1. NPi: A free opensource tool for AI agents to interact with various software tools and applications. Use NPi to build AI agents that can perform tasks such as scheduling meetings, sending emails, and managing GitHub repositories.

calender-negotiator
  1. Ploady: Boost your YouTube video’s performance. Ploady uses AI to generate eye-catching thumbnails, engaging titles, and optimized descriptions, to increase your video’s visibility and reach. Skip the tedious tasks and focus on creating great content.

  2. AgentCrew by MindPal: A collection of AI-powered workflows to automate various tasks such as conducting PESTLE analysis, CV-based interviews, or evaluating IELTS essays. You can use pre-built workflows or customize them to your specific needs. You can even build your own AI agents from the ground up.

  3. Awesome LLM Apps: Build awesome LLM apps using RAG for interacting with data sources like GitHub, Gmail, PDFs, and YouTube videos through natural language conversations. These apps will let you retrieve information, engage in chat, and extract insights directly from content within these platforms.

Hot Takes 🔥

  1. GPT-5 is definitely production ready. That is not a guess.~Flowers from the future

  2. Hint to the AI labs: the first models to work well with structured financial data (especially Excel spreadsheets with formulas) is going to unlock a lot of high-value use cases among firms with a lot of money to spend. ~Ethan Mollick

Meme of the Day 🤡

Context: OpenAI was reportedly holding an event on May 9th where the company was expected to share updates and show product demos. The Information has reported that they have postponed the event.

That’s all for today! See you tomorrow with more such AI-filled content.

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!