🗞️ AI News Recap

Hey TEDchats Fam!

Happy Monday! Get ready for this week's AI news recap! Samsung teases AI innovations at Galaxy Unpacked, Waymo's self-driving car "freaks out" on camera, and Robinhood enhances investing with AI. Plus, Apple joins OpenAI's board, and YouTube introduces new policies for AI-generated content. Dive into all the exciting AI updates right here!

🐇 QUICK BITES — THE PAST WEEK

Samsung's Next Big Reveal: Galaxy Unpacked on July 10th

Samsung's next Galaxy Unpacked event is set for 9AM ET on July 10th in Paris, France, with a clear tagline: "Galaxy AI is here". You can watch the livestream on YouTube here.

Waymo Car 'Freaks Out' During Traffic Stop, Footage Shows

Bodycam footage has been released by Arizona Republic showing a Waymo self-driving car weaving through traffic before pulling into a parking lot on June 19. The footage captures the officer’s perspective as he approaches the empty vehicle. Dispatch records state that the car "FREAKED OUT". According to a Waymo spokesperson, the vehicle encountered inconsistent construction signage and briefly entered an unoccupied oncoming lane for about 30 seconds. Waymo has voluntarily recalled software twice this year following crashes, and the safety of its software is under federal investigation (see Arizona Republic’s released footage).

New Disclosure Policies Released for AI-Generated Election Ads

Google advertisers must now disclose election ads with “synthetic or digitally altered content that inauthentically depicts real or realistic-looking people or events” by selecting the checkbox in the "Altered or synthetic content" section in their campaign settings. Google will show this disclosure in feeds and YouTube Shorts on mobile devices, as well as in-stream ads on phones, computers, TVs, and the web (see Google’s announcement).

Anthropic is Funding AI Tests to "Elevate the Entire Field of AI Safety"

Anthropic is taking significant steps to enhance AI safety by funding new benchmarks designed to evaluate the performance and societal impact of AI models. The initiative, which supports third-party organizations, aims to create rigorous tests that assess AI's capabilities in tasks like cyberattacks, weapon enhancement, and misinformation. With a focus on developing high-quality, safety-relevant evaluations, Anthropic hopes to elevate the entire field of AI safety and set new industry standards (see Anthropic’s announcement).

WhatsApp is Planning an Imaginative AI Experience

WhatsApp's upcoming "Imagine" AI feature, discovered in the latest WhatsApp beta for Android 2.24.14.7 update, lets users create personalized avatars for AI-generated scenarios using Meta's AI Llama model. According to WABetaInfo, who made the discovery, users can simply take a photo and type “Imagine me” in the chat to generate their custom avatars in various imagined settings. While the release date is still uncertain, this feature promises to revolutionize how users interact and express themselves on WhatsApp (see more).

Pixel 9's 'Google AI' Enhances Screenshot Searchability While Maintaining Privacy

Google is introducing a new feature for the upcoming Pixel 9 that promises easier screenshot searches, according to Android Authority. Unlike Microsoft's controversial Recall tool, which captures ‘everything’ using on-device AI, Google's 'Pixel Screenshots' only processes manually taken screenshots. While Recall's rollout was paused due to privacy concerns, Pixel Screenshots allows AI analysis of specific screenshots, enabling content-based searches and bot queries. Users can enable or disable AI processing at any time, maintaining control over their data. This feature is part of a broader AI suite on Pixel 9, including advanced photo editing tools and AI-generated stickers (read more).

Robinhood Acquires Pluto to Boost Investing with AI

Robinhood has acquired AI-powered research platform Pluto to enhance its app with cutting-edge AI tools. The integration of Pluto will enable quicker trend identification, personalized investment strategies, and real-time portfolio optimization for users, marking a significant leap in the app's capabilities (see Robinhood’s announcement).

YouTube Releases a New AI Content Policy

YouTube now allows individuals to request the removal of AI-generated content that mimics their face or voice, expanding their privacy request process. This move aims to protect users from unauthorized synthetic media, particularly concerning in an election year where AI-generated endorsements could influence votes. The platform will assess these requests based on factors like disclosure of AI use, parody context, and public interest. Creators have 48 hours to address complaints, with removals involving complete video takedown and potential blurring of faces, ensuring robust privacy protection (see YouTube’s announcement).

Apple’s App Store Leader is Set to Join OpenAI’s Board

According to a Bloomberg report, Apple has appointed Phil Schiller, head of the App Store, to OpenAI’s governing board as a non-voting observer. This follows the recent partnership to integrate ChatGPT into Apple experiences. Schiller’s role presents opportunities and challenges, particularly with Microsoft also holding a non-voting observer position. While this could complicate strategic discussions, it allows Apple to gain deeper insights into OpenAI’s innovations and ensure its technologies evolve with the latest AI advancements (see Bloomberg’s report).

Cloudflare's New Tool Fights AI Bots

Cloudflare has launched a free tool to protect websites from AI bots scraping data. This addresses growing concerns about AI vendors bypassing protocols like robots.txt, as seen in Perplexity AI’s recent scandal. By improving bot detection models, Cloudflare aims to block these bots effectively, helping site owners safeguard their data (see Cloudflare’s announcement).

Google Releases Its 2024 Environmental Report — But Dodges Future AI Energy Costs?

Google's 2024 Environmental Report highlights green initiatives but avoids disclosing the future energy consumption of its AI operations. Despite showcasing several eco-friendly projects, the company remains vague about AI’s exact energy demands, stating that “predicting the future environmental impact of AI is complex and evolving”. According to the report, Google’s greenhouse gas emissions have surged 48% since 2019, mainly due to increased data center energy consumption and supply chain emissions, casting doubt on its net-zero goal by 2030. Although it may indeed be challenging to predict, this omission raises questions about the true environmental impact of their advanced technologies (see Google’s report).

OpenAI Fixes ChatGPT Mac App's Security Flaw

OpenAI recently addressed a critical security flaw in its ChatGPT macOS app that stored conversations in plain text, making them easily accessible to malicious actors. This vulnerability was discovered by Pedro José Pereira Vieito, who demonstrated how easily another app could read these conversations. OpenAI swiftly released an update to encrypt these locally stored records, enhancing user data protection and ensuring conversations remain secure. This update reflects OpenAI's commitment to maintaining high security standards as their technology evolves (see the demonstration here).

Brazil Halts Meta's AI Training Over Data Privacy Concerns

Brazil’s data protection authority has ordered Meta to stop using Brazilian personal data to train its AI models, citing risks of serious harm to users. Meta faces daily fines of R$50,000 if it fails to comply. This move follows similar actions in the EU, highlighting growing global scrutiny over AI data practices (see the official announcement here).

Meta Makes New Moves to Boost Its Metaverse Platform

Despite selling millions of Quest virtual reality (VR) headsets, Meta has struggled to build a strong user base for its mixed reality platform, facing significant challenges and billions in operating losses. A recent job listing has now revealed Meta's new strategy to integrate generative AI into VR, AR, and mixed reality games, aiming to create immersive, dynamic experiences and lead the evolving AI landscape. Meta also plans to collaborate with third-party creators to enhance game development and content creation (see Meta’s listing here).

🧰 AI TOOLBOX — RECAP

👋 UNTIL NEXT TIME

That's it for this week's AI news recap! Stay tuned for more updates as the world of AI continues to evolve. See you next newsletter for the latest breakthroughs and innovations!