🧠 Meta’s AI Upgrade: What’s New?

Meta AI now supports voice interactions across Messenger, Facebook, WhatsApp, and Instagram, allowing users to get spoken responses and select familiar celebrity voices.

• Users can share and edit photos in chats, with Meta AI recognizing and responding to image content, providing editing tools like background changes and outfit adjustments.

• Testing has begun for automatic translation of Reels, starting with English and Spanish, enhancing content accessibility across languages.

• New tools allow users to generate AI-based images and personalized themes for Facebook and Instagram Stories and feed posts, enhancing customization.

• Meta's business AI tools have expanded, allowing businesses to interact with customers, support commerce, and create AI-powered ads that increase engagement and sales.

🤔 Why It Matters:

Meta AI is leveling up fast, striving to be the most popular AI assistant around, with awesome new features that make your experience even better.

🏃 OpenAI Exodus

OpenAI is losing some big names, like CTO and interim CEO Mira Murati, along with other key figures like Chief Research Officer Bob McGrew and Vice President of Research Barret Zoph.

• Murati’s exit triggered widespread reactions, with employees expressing shock and social media platforms like X filled with memes about the talent drain.

• Several OpenAI co-founders, including Ilya Sutskever, Andrej Karpathy, John Schulman (now at Anthropic), and Daniel Kokotajlo, have also left the company in recent months.

• Elon Musk, former OpenAI co-founder, has reignited a lawsuit against the company, accusing it of misleading him about its non-profit status.

🤔 Why It Matters:

Despite the wave of high-profile exits, OpenAI remains a key player in the AI industry, with products like ChatGPT, GPT-4, and DALL-E leading the market. However, with rising competition from Meta AI, the company’s long-term dominance could be challenged. OpenAI’s future is in question, with ongoing negotiations for a $150 billion valuation funding round involving investors like Apple and Nvidia. Additionally, a shift away from its non-profit structure and a potential payout to CEO Sam Altman could signal further changes ahead.

🤖 A Game-Changer in AI-Powered Chip Design

Google introduced AlphaChip, an AI-based reinforcement learning tool designed to accelerate and optimize chip layouts, significantly reducing time and cost in chip design.

• AlphaChip can complete complex chip floorplans, which traditionally take 24 months by human designers, in just a few hours. It improves chip performance, power efficiency, and reduces wire length across various processors, including Google’s Tensor Processing Units (TPUs).

• AlphaChip has already been used in designing TPUs and MediaTek’s Dimensity 5G chips, with future applications across more processor types.

• AlphaChip views chip layout design as a game, using a graph neural network to improve its ability to place circuit components on a grid efficiently.

🤔 Why It Matters:

By automating the most labor-intensive phase of chip development, AlphaChip democratizes AI-assisted design, potentially reducing development times and costs for the entire semiconductor industry. AlphaChip has the potential to revolutionize chip manufacturing. It could lead to faster, smaller, and more energy-efficient processors in all industries.

⚔️ Cerebras Takes On Nvidia With New IPO Filing

AI chipmaker Cerebras Systems has filed for an initial public offering (IPO) under the ticker symbol “CBRS” on the Nasdaq.

• Cerebras competes with major players like Nvidia, boasting a chip (WSE-3) with more cores and memory than Nvidia’s H100, and also offers cloud-based services.

• Cerebras reported a net loss of $66.6 million on $136.4 million in sales for the first six months of 2024. Operating expenses increased due to personnel costs to support growth.

• AI chips are a competitive space with major cloud providers like Amazon, Google, and Microsoft developing their own solutions. In 2023, Group 42 accounted for 83% of Cerebras’s revenue.

🤔 Why It Matters:

With its IPO, Cerebras is positioning itself as a strong competitor to Nvidia, leveraging its advanced chip technology and strategic partnerships like the $1.43 billion commitment from Group 42. However, supply chain risks and a crowded AI chip market pose significant challenges as Cerebras seeks to expand its market share.

📰 California Governor Halts AI Safety Bill

Governor Gavin Newsom vetoed the Safe and Secure Innovation for Frontier AI Models Act (SB 1047), citing concerns over its broad application and potential to stifle innovation in California’s AI sector. Newsom argued that the bill would impose burdensome regulations on both basic and advanced AI models, creating a false sense of security.

Read More →

📰 Meta’s AI Posts

Meta announced it will test AI-generated posts in Facebook and Instagram feeds, tailored to users’ interests and even featuring their likeness. Users can engage with or opt out of this “imagined for you” content. This move builds on Meta’s use of algorithms to surface content users might enjoy, pushing beyond posts shared by friends to personalized, AI-generated images and scenes.

Read More →

📰 UC Berkeley Introduces DocETL

UC Berkeley researchers have introduced DocETL, an open-source, low-code AI tool designed to process unstructured documents using large language models (LLMs). DocETL enables summarization, classification, question-answering, and entity resolution on unstructured data via a declarative YAML interface, making it accessible for non-experts.

Read More →

📰 OpenAI Faces Financial Strain

OpenAI’s monthly revenue hit $300 million in August 2024, a 1,700% increase since early 2023. The company expects $3.7 billion in annual sales this year, with projections for $11.6 billion in 2025. Despite high revenue, OpenAI is projected to lose $5 billion this year due to operating costs. Expenses are growing in tandem with user base expansion, with 350 million monthly users across services like ChatGPT, a 250% increase since March 2024.

Read More →

📰 Meta Won’t Say If Smart Glasses Images Train AI

Meta’s Ray-Ban Meta glasses, equipped with AI-powered cameras, capture photos both intentionally and passively. Meta, however, will not disclose whether these images are used to train AI models. When asked directly, Meta representatives declined to confirm or deny whether the photos from smart glasses users are used for AI training, raising concerns about user privacy.

Read More →

📰 OpenAI Academy to Support AI Innovation in Developing Nations

OpenAI Academy wants to support developers and organizations in poorer countries to use AI for community problem-solving and economic growth. The Academy will provide AI technology, training, and access to $1 million in API credits, along with technical guidance and a global developer network for collaboration and innovation.

Read More →

💡 SupaSend

SupaSend is an awesome iOS app that lets you easily capture and send your ideas, notes, or thoughts to different “second brain” apps. It works with lots of email clients and productivity apps, like Evernote, Notion, and Reflect, so you can organize your thoughts wherever you are.

Check Now →

🔎 Canyon

Canyon is a platform that makes job hunting and applying easier. It’s got features like an AI-powered Resume Builder, Resume Score to optimize it, and an autofill tool to speed up job applications. It also has tools to help you keep track of applications, resources to prepare for interviews with personalized tips, and a browser extension for easily applying to jobs and keeping track of them.

Check Now →

🎥 Whatmore Studio

Whatmore Studio helps e-commerce brands effortlessly make pro marketing videos with AI. It takes product images and URLs and turns them into dynamic videos with music and captions.

Check Now →