🛎️The Price of Choice
Good Morning, AI Enthusiasts!
Silicon Valley's Game of Thrones is back with a new season. When you pick a side, you also pick an enemy.
SPONSORED BY
QUICK HITS
- Bellingcat's recent analysis reveals that ChatGPT outperformed Google Lens in geolocating images, marking significant advancements in AI's geolocation capabilities.
- AMD unveiled its Helios rack-scale systems, featuring 72 MI400 GPUs and 2.9 exaFLOPS performance, to compete with Nvidia's upcoming Vera Rubin NVL144.
- Taiwan adds Huawei and SMIC to its export control list, aligning with U.S. efforts against China's AI ambitions.
- Barksdale Federal Credit Union has successfully implemented Scienaptic AI's credit underwriting platform, enhancing efficiency and approval rates.
- Concerns arise over a potential mass exit of Indian tech workers from the US, fueled by Reddit buzz.
DATA
Meta's $14B Gamble Sparks Industry Ban on Scale AI
The AI industry is no stranger to drama, but few stories have sent as many shockwaves through Silicon Valley as Meta’s jaw-dropping $14 billion acquisition of a 49% stake in Scale AI. This high-stakes move has not only raised eyebrows but has also triggered a cascade of reactions from tech giants and competitors alike.
- Following the acquisition, Google yanking a $200 million contract and industry heavyweights like Microsoft, xAI, and OpenAI suddenly ghosting Scale AI, the plot thickens.
- The industry’s reaction underscores the paramount importance of trust; companies are wary of sharing data with a stakeholder like Meta, fearing potential competitive advantages.
- Competitors like Labelbox are benefitting from the situation, as companies increasingly prioritize neutrality in their data partnerships to avoid conflicts of interest.
In a landscape where trust is paramount, the rapid withdrawal of major players from Scale AI highlights a critical lesson: data sovereignty and neutrality are no longer optional; they are essential for survival.
TOGETHER WITH MACROCOSMOS
Pretraining is the hardest part of AI to decentralize - until now
Vast data centres, enormous compute, and insatiable energy demands: only big tech can train LLMs. Even open-source models like Llama or Qwen are centrally trained, leaving DeAI dependent on centralized competitors.
Macrocosmos, an open-source AI lab on Bittensor, just launched IOTA - Incentivized Orchestrated Training Architecture. It allows anyone, anywhere to participate in LLM training-at-home:
✅ No memory constraints = wider accessibility.
✅ Fair, real-time rewards = co-ownership with proportionate attribution.
✅ 128× compression + trustless weight merging = global scale, fast.
Check out our interactive 3D dashboard. Join now to contribute compute, get rewarded, and train LLMs - and help make data centres a thing of the past.
AI VIDEO
Midjourney Launched First Video Model
Midjourney, the company renowned for its AI-generated art tools, unleashed its very first video model into the wild. The internet, as expected, did not disappoint: Reddit exploded with nearly 3k likes in a matter of hours.
- Midjourney's transition from still images to video is monumental, requiring the ability to create a series of coherent and dynamic frames rather than just a single compelling image.
- The video model uses a sophisticated transformer architecture, allowing it to capture long-range dependencies in video, ensuring that actions flow smoothly across frames.
- The model generates clips at resolutions up to 720p and frame rates between 12 and 24 frames per second, making it suitable for short-form content.
- The model retains Midjourney's signature aesthetic, allowing users to specify styles while maintaining character consistency across frames, a key challenge in AI video generation.
As we embrace this new era of video generation, we can expect an explosion of creativity, innovation, and perhaps even a few viral sensations—like that cyberpunk cat on a hoverboard. With every new development, we edge closer to a future where anyone can turn their wildest ideas into visual reality, reshaping the landscape of digital art forever.
CRM
Salesforce's CRMArena-Pro Benchmark: LLM Agents Struggle to Pass the CRM Test
In the realm of CRM, Salesforce has unveiled a new benchmark, CRMArena-Pro, which critically assesses the capabilities of LLM-based AI agents. This initiative sheds light on the current limitations of AI in handling complex, real-world CRM scenarios, suggesting that these technologies have not yet reached the level of competence required to fully replace human professionals in this field.
- Salesforce's CRMArena-Pro benchmark is a rigorous testing platform designed to evaluate LLM agents on real-world CRM tasks, moving beyond simplistic interactions to encompass multi-turn dialogues and complex business activities.
- The benchmark revealed that advanced LLM agents, such as Gemini 2.5 Pro, achieved a mere 58% success rate on single-turn tasks and a disappointing 35% on multi-turn dialogues, underscoring significant deficiencies in handling nuanced conversations.
- The findings indicate that while LLM agents can assist with simple tasks, their inadequacy in complex CRM scenarios and issues with data confidentiality make them unsuitable for full-scale deployment in enterprises.
The revelations from Salesforce's CRMArena-Pro benchmark serve as a crucial wake-up call for businesses relying on AI in their CRM efforts. The stark reality is that despite the impressive advancements in AI technology, these agents are not yet equipped to manage the intricate and often unpredictable demands of real-world customer interactions.
TRENDING TOOLS
- 🎙 VoiceType is an AI-powered speech-t0-text app that boosts writing speed by 9 times with 99.7% accuracy.
- 🧐 Readefine rephrases the internet.
- 🖼️ Pixlr is a free online AI image generator.
- 💬 OnlyBots is the world's first social network exclusively for AI bots.
- 📀 Suno-Top is a free downloader for Suno AI-generated music.
- 🪶 CopyOwl is the First AI Research Agent, deep research on any topic in one click. (Featured)
- 📚 Heardly is the Fast Way to read Best Book. (Featured)
- 🦾 Flot AI writes, reads, and remembers across any apps and webs. (Featured)
1 Million+ AI enthusiasts are eager to learn about your product.
AI Secret is the world’s #1 AI Newsletter, boasting over 1 million readers from leading companies such as OpenAI, Google, Meta, and Microsoft. We've assisted in promoting Over 500 AI-Related Products. Will yours be the next?
What We Can Offer:
- Launch an Advertising Campaign
- Introduce New Product or Features
- Other Business Cooperation
Email our co-founder Mark directly at mark@aisecret.us if the button fails.
Magna
Magna
Magna
