🗞 OpenAI Shakes Up Safety Oversight, Altman Steps Down 🔄

🔎 The Latest on the AI Frontier:

• 🔄 OpenAI Reshuffles Safety Committee, Altman Steps Down

• 🍓 OpenAI's 'Strawberry' Model: Advanced yet Concerning

• 💰 Salesforce Doubles Down on AI with $1B Investment Fund

• 🏷️ Google to Introduce AI-Generated Image Labels in Search

• 📦 Amazon Harnesses AI to Revolutionize Same-Day Delivery

• 🤖 Microsoft Launches Copilot Pages for Team Collaboration

• 💡 Gates Endorses AI Startups as Prime Opportunity for Entrepreneurs

• 🔋 iOS 18 Update Sparks Battery Drain Complaints

• 🗳️ Congress Lags on AI Election Safeguards for 2024 Vote

• 👀 Other news you might find interesting

🔄 OpenAI CEO Sam Altman steps down from safety committee amid transparency push. Link

• OpenAI restructures its safety and security committee to be fully independent, removing CEO Sam Altman.

• New committee chair Zico Kolter will lead efforts to review major AI models and potentially delay releases.

• Move aims to address critics' concerns and increase transparency in OpenAI's development process.

🍓 OpenAI's new 'Strawberry' model raises AI safety concerns. Link

• The o1 model (nicknamed Strawberry) shows advanced reasoning capabilities but also potential for deception and weapons assistance.

• OpenAI rates Strawberry as "medium" risk for nuclear, biological, and chemical weapons - a first for the company's public models.

• The release highlights the safety paradox in AI development: more advanced models may be both more dangerous and easier to monitor.

💼 Salesforce Ventures doubles AI fund to $1 billion, reflecting strong corporate interest in AI. Link

• The new $500 million AI fund adds to last year's $500 million commitment, showcasing Salesforce's growing focus on AI investments.

• AI startups raised $4.3 billion in August, accounting for 24% of total venture funding despite overall market slowdown.

• Salesforce Ventures has participated in 10 AI-related deals recently, including investments in Ensemble AI and Protect AI.

🖼️ Google to flag AI-generated images in search results. Link

• New feature will use C2PA authentication standard to identify if photos were taken by camera, edited, or AI-generated.

• Google Search and ads to adopt C2PA metadata, with plans to integrate into YouTube later this year.

• Initiative aims to combat misinformation, but faces challenges in widespread adoption across hardware and software platforms.

🚚 Amazon leverages AI to boost same-day shipping capabilities. Link

• Generative AI optimizes delivery routes, improves warehouse robotics, and predicts inventory placement.

• 60% of Prime orders in top 60 U.S. metro areas delivered same or next day in March.

• Amazon's warehouse robots increased from 350,000 in 2021 to over 750,000 in 2023, now using AI for smarter navigation.

🤝 Microsoft unveils Copilot Pages for collaborative AI workspace. Link

• New feature allows teams to work together with AI in real-time, combining web and work data into a shared canvas.

• Copilot Pages integrates with Microsoft 365 and will be available to all subscribers later this month.

• Microsoft also launches Copilot agents for businesses, acting as virtual employees to automate tasks and respond to queries in Teams and Outlook.

💡 Bill Gates: AI startups are the 'fantastic opportunity' for today's entrepreneurs. Link

• Gates suggests he'd start an "AI-centric" company if launching Microsoft today, citing massive funding potential.

• He advises focusing on niche AI applications to stand out in the crowded market, which has seen $26 billion in investments this year.

• Gates encourages young entrepreneurs to look for ways AI can improve businesses or daily life, calling it "the frontier" for innovation.

🔋 iOS 18 users report battery drain issues after update. Link

• Some iPhone users complain of rapid battery depletion after installing the new iOS 18.

• Experts say this is normal after major updates due to background processes and recalibration.

• Tips to mitigate: Reboot device, update apps, reduce screen brightness, and give it time to stabilize.

🗳️ Congress unlikely to pass AI election safeguards before 2024 vote, despite urgent warnings. Link

• Despite bipartisan support and multiple bills introduced, time has run out for meaningful AI regulation to protect the 2024 elections.

• Sen. Schumer's handpicked AI team has made little progress, with key bills failing to pass unanimous consent in July.

• In the absence of federal action, 19 states have passed their own laws to limit deepfakes in elections, while federal agencies like the FCC are using existing rules to combat AI misuse.

More news you might find interesting:

• 💸 Meta launches AI Accelerator TLV for Israeli startups. Link

• 🚫 Meta bans RT and Russian state media networks globally. Link

• 👤 IG introduces 'teen accounts' with enhanced parental controls. Link

•  💨 SambaNova challenges OpenAI with high-speed AI demo. Link

• 🚀 Perplexity integrates OpenAI's o1 model, expands capabilities. Link

• 🧠 AI meets fruit fly brain, revolutionizing neural network modeling. Link

• 🇨🇳 China races to develop homegrown AI chip alternatives to Nvidia. Link

• 🧬 Rakovina Therapeutics partners with Variational AI to accelerate cancer drug discovery. Link

• 🌍 Microsoft boosts AI investment in Middle East, partnering with Abu Dhabi's G42. Link

• 🤖 Slack pivots to become AI agent hub, raising questions about user needs and privacy. Link

Enjoying the newsletter? Please consider supporting us:

Have any feedback? Send us an email