🗞️ AI News Recap

Hey TEDchats Fam!

Happy Monday! “AI is the new electricity”, said Andrew Ng, a leading AI researcher and co-founder of Google Brain. This week’s AI news encapsulates this sentiment perfectly, from groundbreaking partnerships and legal battles to innovative tools reshaping industries. Dive into the latest developments and see how AI continues to revolutionize the world around us.

We have partnered with Vanta to bring you today’s newsletter. Discover the latest in AI security and snag your spot for their live session → here

🐇 QUICK BITES — THE PAST WEEK

Another Major News Organization Partners with OpenAI!

TIME and OpenAI have entered a multi-year content agreement and strategic partnership to bring TIME's reputable journalism to OpenAI's products, including ChatGPT. This collaboration allows OpenAI to utilize TIME's extensive archives, spanning 101 years, to enrich its offerings. TIME's content will be used to answer user queries, complete with proper citations and links back to the original articles. This action follows similar agreements made by other media companies to license content to OpenAI for training its AI models, including News Corp, which provides access to archives from major outlets like The Wall Street Journal and The New York Post, as well as The Financial Times, Vox Media, Axel Springer, and The Associated Press (see OpenAI’s announcement).

Meanwhile, Another News Organization Sues OpenAI (and Microsoft)

The Center for Investigative Reporting (CIR) has filed a lawsuit against OpenAI and Microsoft, accusing them of using CIR journalists' creative works without compensation or authorization to train their AI models. CIR seeks actual damages and defendants' profits, or statutory damages of at least $750 per infringed work and $2,500 per Digital Millennium Copyright Act (DMCA) violation. This lawsuit aligns CIR with other media organizations, such as The New York Times, Chicago Tribune, New York Daily News, Raw Story, The Denver Post, AlterNet, and The Intercept, which have taken similar legal actions (see CIR's lawsuit here).

Consent-First Approach: YouTube's Plan to Launch New AI Music Tools

Last year, YouTube introduced Dream Track, an AI feature that creates music in the style of artists like Charlie Puth and Sia, initially supported by ten artists. The main challenge is convincing major record labels to participate, as they oppose unauthorized use of their content and have even sued AI music companies for copyright breaches, among other actions (see recent controversial examples here, here, and here). Despite this, YouTube is reportedly negotiating and offering substantial lump-sum payments to license songs for AI training. They aim to attract artists and launch new AI tools later this year, following a “consent-first” philosophy to address copyright and artist rights (read more).

Toys 'R' Us Resurrected with AI Magic

The beloved toy store brand, Toys 'R' Us, has made an unexpected comeback with an AI-generated video using OpenAI's new tool, Sora. Despite past decline, this effort suggests a revival, fueled by the media frenzy over the video's flaws. Embracing the "no such thing as bad publicity" mantra, Toys 'R' Us plans new stores and innovative digital content (watch the video here).

Formation Bio Raises $372M to Revolutionize Drug Development with AI

Formation Bio, a startup leveraging AI for drug development, has raised an impressive $372 million in a Series D round led by Andreessen Horowitz. With a total of over $600 million raised, Formation is set to enhance clinical trials and streamline drug R&D processes, aiming to accelerate the journey from discovery to market. Notably, the company collaborates with OpenAI and Sanofi, harnessing AI to predict drug efficacy and toxicity, with its eczema treatment already in phase 3 trials (see Formation Bio’s announcement).

Google is Rolling Out 110 New Languages to Google Translate

Google Translate is set to nearly double its language support, adding 110 new languages, thanks to advancements with Google's PaLM 2 AI language model. This expansion includes long-requested languages like Cantonese, which poses unique challenges due to its overlap with Mandarin in writing. Additionally, about a quarter of the new languages come from Africa, reflecting Google's commitment to inclusivity. With this update, Google Translate will support over 200 languages, catering to millions more users worldwide (see Google’s announcement).

Firefox Nightly Introduces Customizable AI Chatbot Options

Mozilla's Firefox Nightly now lets you choose your preferred AI chatbot. Options include ChatGPT, Google Gemini, HuggingChat, and Le Chat Mistral, providing a customizable AI experience tailored to your needs. This experimental opt-in feature aims to refine user preferences before a wider release, allowing users to “summarize information, simplify language, or test their knowledge, all without leaving their current web page” (see Firefox’s announcement).

Gmail Gets Smarter with AI: Introducing Gemini Sidebar

Google is revolutionizing email management with the rollout of Gemini AI in the Gmail side panel. This new feature, available to paid subscribers, can summarize email threads and draft responses with ease. Leveraging the powerful Gemini 1.5 Pro model, it aims to streamline your inbox experience, though users are advised to double-check AI-generated content for accuracy. In addition to Gmail, Google is integrating Gemini AI into Docs, Sheets, Slides, and Drive, promising a more efficient workflow across its suite of productivity tools. These advancements highlight Google’s commitment to enhancing user productivity through cutting-edge AI technology (see Google’s announcements here and here).

Meta’s AI Labeling Strategy is Misfiring!

Meta is under fire from photographers for incorrectly labeling real photos as "Made by AI". The issue arose when images, even with minor edits using tools like Adobe's cropping and Generative Fill, were marked as AI-generated. For instance, former White House photographer Pete Souza's image of a basketball game and a photo of the Kolkata Knight Riders winning the IPL were mislabeled. This labeling only appears on mobile devices, not on the web, complicating the situation further. Photographers argue that such a broad application of the label diminishes its meaning and relevance. In response, Meta is reviewing its labeling criteria to better reflect the actual use of AI in photos. You can find Meta’s AI labeling process here and a summary of complaints by The Verge here.

ChatGPT Now Available for All macOS Users

Initially launched at Apple’s Worldwide Developer Conference in June, OpenAI and Apple's collaboration integrates ChatGPT across iPhone, iPad, and Mac (see Apple’s announcement). After starting with Plus subscribers, the ChatGPT desktop app is now widely available for macOS 14+ on M1 or newer chips. Users can access ChatGPT with Option + Space and enjoy features like screenshots and file uploads. Later this year, deeper integration with Siri and a Windows version of the app are expected (see OpenAI’s announcement).

Stability AI Secures Crucial Funding Amid Financial Turmoil

Stability AI, the creator of the groundbreaking Stable Diffusion model known for its revolutionary advancements in image generation and AI art, has received a significant cash injection from high-profile investors, including Sean Parker (Facebook’s former president) and Eric Schmidt (Google’s former CEO). This crucial funding comes as Stability AI battles a severe cash crunch, with unpaid cloud bills and declining valuation (see Stability AI’s announcement).

Reddit is Cracking Down on AI Crawlers

Reddit is updating its Robots Exclusion Protocol (robots.txt file) to prevent AI companies from scraping its content for training models. This move, which includes rate-limiting and blocking unauthorized bots, aims to protect Reddit’s data integrity, especially from companies lacking agreements. Earlier this year, Reddit secured data licensing deals worth $203 million, as stated in their IPO prospectus, highlighting their data’s role in training large language models (LLMs) and monetization. The modifications to the robots.txt file underscore Reddit's commitment to protecting user content and impose strict conditions on AI developers who access data either in bad faith or without a monetization agreement (see Reddit’s announcement).

ChatGPT's “Advanced Voice Mode”, Capable of Understanding and Responding to Emotions, is Delayed

OpenAI’s highly anticipated “advanced voice mode” for ChatGPT, initially expected to roll out in late June, has been postponed to July due to ongoing improvements and scaling preparations. The company is focused on regulating content and refining the user experience to ensure seamless, real-time interactions. Although the Voice Mode launch for ChatGPT Plus customers might be delayed until the fall, OpenAI still plans to roll out new video and screen-sharing capabilities. These new features will enable ChatGPT to solve math problems from images and explain device settings, enhancing functionality across both mobile and desktop platforms (see OpenAI’s announcement).

Microsoft’s Mustafa Suleyman Expresses Admiration for OpenAI’s Sam Altman

In a recent interview at the Aspen Ideas Festival, Mustafa Suleyman, CEO of Microsoft AI, expressed his admiration for OpenAI CEO Sam Altman. Suleyman highlighted his trust in Altman's sincere commitment to AI safety and his belief in genuine regulation. Despite the intense competition between their companies, Suleyman underscored the importance of their partnership and the constructive dialogue on AI's future. He noted that taking the time to regulate and ensure the safe development of AI technologies is crucial, even if it means slowing down progress temporarily (watch the full interview here).

SoftBank Partners with Tempus AI to Bring AI-Driven Medical Services

SoftBank Group Corp. has launched a joint venture with Chicago-based Tempus AI to bring AI-powered personalized medical services to Japan, starting with oncology. This collaboration aims to revolutionize cancer research using data analysis and AI insights, with a goal to offer tailored treatment plans for patients (see Softbank’s announcement). Tempus, connected to around 50% of U.S. oncologists, also recently received U.S. FDA clearance for its AI-based ECG-AF device, which identifies patients at risk of atrial fibrillation/flutter (see Tempus AI’s announcement).

Major Record Labels Sue AI Music Companies for Copyright Violations

In a landmark action, major record labels such as Universal Music Group, Sony Music Entertainment, and Warner Records Inc. have initiated lawsuits against AI music companies Suno and Udio for significant copyright breaches. These AI companies allegedly used artists' works without permission to produce popular tracks and are now facing claims for damages that could reach up to $150,000 per infringed work (see Suno's case here and Udio's case here).

OpenAI Unveils CriticGPT — GPT-4's Sidekick

Say hello to CriticGPT, OpenAI’s latest innovation designed to hunt down bugs in GPT-4’s code. This bug-busting sidekick is joining the RLHF (Reinforcement Learning from Human Feedback) alignment pipeline, helping humans keep a closer eye on AI (see OpenAI’s announcement).

🧰 AI TOOLBOX — RECAP

Join the live session: automate compliance & streamline security reviews

Whether you’re starting or scaling your company’s security program, demonstrating top-notch security practices and establishing trust is more important than ever.

Vanta automates compliance for SOC 2, ISO 27001, and more, saving you time and money — while helping you build customer trust.

And, you can streamline security reviews by automating questionnaires and demonstrating your security posture with a customer-facing Trust Center, all powered by Vanta AI.

Join Vanta’s 45-minute live session on July 9th at 12 pm PST to see the platform in action and ask your questions.

👋 UNTIL NEXT TIME

As we’ve seen, AI is rapidly transforming diverse sectors, echoing Andrew Ng's comparison to electricity. From new collaborations and legal challenges to groundbreaking innovations, the AI landscape is evolving at a remarkable pace. Stay tuned for more updates as we continue to witness the powerful impact of AI on our world.

In partnership with