Meta’s “Open” Llama Models Come With a Revenue-Sharing Catch

✨TodayOnAI’s Daily Drop

• Meta’s “Open” Llama Models Come With a Revenue-Sharing Catch

• FuriosaAI Rejects $800M Meta Acquisition to Stay Independent

• OpenAI Upgrades ChatGPT Voice Mode to Reduce Interruptions

• DeepSeek’s Viral AI Models Disrupt Global Market, Trigger U.S. Backlash

• 💬 Let’s Fix This Prompt

• 🧰 Today’s AI Toolbox Pick

META

Meta’s “Open” Llama Models Come With a Revenue-Sharing Catch

🚀 TodayOnAI Insight: Despite public claims that monetization isn’t its goal, Meta is generating revenue from its open-source Llama models through partnerships with major cloud providers—raising new questions amid an ongoing copyright lawsuit.

🔍 Key Takeaways:

• Meta shares revenue from Llama models hosted by third-party platforms like AWS, Google Cloud, and Snowflake, according to a newly unsealed court filing in Kadrey v. Meta.

• While developers can self-host Llama models for free, partners provide additional services that simplify deployment.

• Mark Zuckerberg previously hinted at licensing and monetization strategies, including AI-enhanced business messaging and ads.

• The lawsuit alleges Meta trained Llama on pirated e-books and may have facilitated further infringement via torrenting.

• Meta plans to double its capital expenditures to $60–$80 billion in 2025, largely to scale AI infrastructure and teams; a paid tier for Meta AI is reportedly in development.

💡 Why This Stands Out: Meta’s “open” AI strategy is more commercially entangled than it appears, revealing a pragmatic approach to offsetting massive AI investments. As legal scrutiny intensifies, the blurred lines between openness and monetization could reshape how major tech firms position their AI ecosystems. Can truly open AI coexist with billion-dollar business models?

FURIOSA

FuriosaAI Rejects $800M Meta Acquisition to Stay Independent

🚀 TodayOnAI Insight: AI chip startup FuriosaAI has reportedly turned down an $800 million acquisition offer from Meta, choosing to remain independent and push ahead with its next-gen chip roadmap—a strategic move that underscores the growing power of specialized AI hardware players.

🔍 Key Takeaways:

• FuriosaAI declined Meta’s acquisition offer due to disagreements over post-acquisition strategy and organizational fit, not valuation.

• The South Korean startup is finalizing a $48M funding round and preparing to launch its second AI chip, Renegade (RNGD), later this year.

• RNGD chips are optimized for reasoning tasks and have already been tested with partners like LG AI Research and Aramco.

• Meta has been aggressively investing in custom AI chips and infrastructure to reduce its reliance on Nvidia, with up to $65B in AI spending planned for 2025.

• Founded by ex-Samsung and AMD engineer June Paik, FuriosaAI is positioning itself as a competitive alternative to Nvidia and AMD in the AI chip race.

💡 Why This Stands Out: FuriosaAI’s decision to stay independent signals growing confidence among upstart chipmakers in challenging incumbent giants. As demand for AI-optimized hardware surges, the balance of power is shifting—will more startups reject Big Tech’s advances in favor of long-term vision and control?

OPENAI

OpenAI Upgrades ChatGPT Voice Mode to Reduce Interruptions

🚀 TodayOnAI Insight: OpenAI has updated ChatGPT’s Advanced Voice Mode to improve conversational flow and reduce interruptions, aiming to make the AI assistant feel more natural and personable—just as competition in the voice assistant space heats up.

🔍 Key Takeaways:

• New update prevents the AI from interrupting users during natural pauses, such as when thinking or breathing.

• Available to both free and paid ChatGPT users, with subscribers gaining access to an even more polished assistant personality.

• Voice responses are now described as “more direct, engaging, concise, specific, and creative,” per OpenAI.

• The update was announced by post-training researcher Manuka Stratta via OpenAI’s official channels.

• Comes as startups like Sesame (backed by a16z) and tech giants like Amazon push forward with LLM-powered voice assistants.

💡 Why This Stands Out:
OpenAI’s voice assistant is evolving beyond scripted exchanges toward real-time, fluid conversation—an essential leap as voice becomes a primary interface for AI. With rivals gaining traction, the race is on to make AI not just smart, but truly listenable. Will OpenAI’s head start in voice hold up against fast-moving challengers?

DEEPSEEK

DeepSeek’s Viral AI Models Disrupt Global Market, Trigger U.S. Backlash

🚀 TodayOnAI Insight: Chinese AI lab DeepSeek has exploded onto the global scene, topping app store charts and shaking investor confidence in U.S. AI dominance. With compute-efficient models and viral adoption, DeepSeek is forcing a reevaluation of the AI arms race—politically and technically.

🔍 Key Takeaways:

• DeepSeek, backed by hedge fund High-Flyer Capital, began in 2023 and quickly launched competitive models like DeepSeek-V3 and the R1 reasoning model.

• R1 performs on par with OpenAI’s o1 in internal benchmarks and excels in math and science tasks due to its self-checking reasoning approach.

• Despite U.S. export restrictions, DeepSeek trained its models on Nvidia’s H800 chips and built in-house data centers.

• DeepSeek offers its models under commercially permissive licenses and has avoided VC funding while pricing its offerings below market norms.

• Its rapid rise has triggered backlash: the U.S., New York state, South Korea, and other entities have banned DeepSeek on government devices, citing political and regulatory concerns.

💡 Why This Stands Out: DeepSeek’s ascent signals more than technical prowess—it highlights how geopolitics, pricing strategy, and compute access are redefining AI competition. Its disruptive presence is reshaping global AI economics and intensifying scrutiny of foreign-developed models. As reasoning models gain traction, are we seeing the next paradigm shift in AI performance—and power?

✨ See how a simple prompt upgrade can unlock better AI output.

🔹 The Original Prompt

At first glance, this prompt might seem okay. But it's too broad — and that limits the quality of AI-generated results. Let’s improve it using prompt engineering best practices.

✅ The Improved Prompt

💡 Why It's Better

• Specific audience: Targets decision-makers in mid-sized companies.

• Contextual focus: Emphasizes emerging tech and practical solutions.

• Actionable output: Requests summaries and titles to spark execution.

• Tone and style: Guides the type of content (insightful, engaging, relevant).

🛠️ Learn how to adapt this prompt for SaaS, AI tools, dev teams & more →
Read the full PromptPilot breakdown

💡 Bonus Tool: Want to generate and master prompts instantly?
👉 Try PromptPilot by TodayOnAI (Free to use)

📰 More from the AI World

• Copilot Makes Discovering Ideas Feel Like a Conversation

• Vevo & Arc Institute Release 300M-Cell Atlas to Advance Drug Discovery with AI

• Meta Launches Aria Gen 2 to Power the Future of Perception & Contextual AI.

• Talk to Perplexity: Real-Time Voice Answers Now on iOS

🧰 Today’s AI Toolbox Pick

• 🍋LemonSqueezy (Finance Tool): Handles the tax compliance burden so you can focus on more revenue with less headache.

• 💻ZipWP (Web Design Tool): Creates stunning websites in seconds.

• ⚙️DupDub (Content Tool): An all-in-one content creation platform that allows you to craft your content effortlessly and streamline your workflow.