Amazon’s new AI model

Plus, 👀 Best AI Video Generators and Mercury: The AI Model That’s 10x Faster Than Traditional LLMs


Hello there! Ready to dive into another upgrading, Mind-boggling, and Value-filled Shift?

If only this was true 🥲

Today we have:

🤖 Amazon Enters the AI Reasoning Race With Nova Model

👀 Best AI Video Generators

🚀 Mercury: The AI Model That’s 10x Faster Than Traditional LLMs

🛠️ Tools and Shifts You Cannot Miss

🤖 Amazon Enters the AI Reasoning Race With Nova Model
Insights from Techcrunch

Amazon is developing an advanced reasoning AI model under its Nova brand, set for a June 2025 release. This marks Amazon’s biggest push into AI reasoning, putting it in direct competition with OpenAI, Anthropic, and Google.

The Decode:

  • Hybrid Reasoning for Smarter AI - Amazon’s model will blend fast response generation with complex multi-step problem-solving, similar to Anthropic’s Claude 3.7 Sonnet.

  • Competing on Cost & Performance - Amazon reportedly plans to undercut competitors on pricing, despite the challenge posed by DeepSeek’s ultra-low-cost models. The goal is to rank among the top five AI models, delivering affordable yet high-quality performance.

  • A Strategic Shift in Amazon’s AI Playbook - Despite an $8 billion investment in Anthropic, Amazon is aggressively developing its own AI infrastructure, signaling a move towards independence in the AI space. With an upgraded AI-powered Alexa+ also in the works, Amazon is positioning itself as a serious contender in generative AI.

Amazon is no longer just backing AI—it’s building its own. With cost efficiency, reasoning capabilities, and deep AI investments, Amazon is gearing up to compete at the highest level, challenging both its partners and rivals in the AI race.


👀 Best AI Video Generators 
Insights from Barsee via X

🚀 Mercury: The AI Model That’s 10x Faster Than Traditional LLMs
Insights from Inception

Inception Labs has introduced Mercury, a groundbreaking diffusion large language model (dLLM) that generates text all at once rather than word-by-word like traditional LLMs. This new approach, similar to how image generators like Midjourney and DALL-E work, enables faster, more accurate responses.

The Decode:

  • Diffusion Over Autoregressive Models - Unlike ChatGPT, Claude, or Gemini, which generate text sequentially, Mercury creates an entire response at once and refines it iteratively.

  • Blazing-Fast Speed & High Performance - Mercury delivers over 1000 tokens per second on standard NVIDIA H100s, making it up to 10x faster than traditional models like GPT-4o Mini and Claude 3.5 Haiku. Developers prefer Mercury’s code completions, with tests showing it outperforms larger models while maintaining superior speed.

  • A Game-Changer for Code & AI Applications - Mercury Coder, the first publicly available dLLM, is optimized for code generation and has already surpassed GPT-4o Mini and Gemini-1.5-Flash in Copilot Arena benchmarks. The diffusion-based approach also lowers inference costs, making AI-powered automation, customer support, and enterprise applications more efficient.

Diffusion models could reshape AI text generation, offering faster, smarter, and more efficient outputs. With early adopters already seeing benefits, the question is: how long until OpenAI, Google, or Microsoft release their own version?


🛠️  AI Tools for the Shift

⚖️ Cardamon – Automate regulatory compliance with AI and launch financial products 10X faster. Get from regulations to requirements in minutes.

Chirp – The ultimate Apple Watch assistant for seamless voice-to-text, smart searches, and instant messaging—all from your wrist.

📄 Breyta.ai – Instantly synthesize and extract insights from multiple unstructured data files, backed by evidence.

🛠️ ToolJet – Describe your app, and ToolJet builds it—UI, database, queries, everything. The fastest way to build internal tools.

🧠 Comigo – AI-driven ADHD support that combines therapy with practical execution for better daily management.


📄Quick Shifts

❓ Want Done-for-You UGC Ads in 14 Days? With Insense’s managed service, brands like Opopop and Bones Coffee get high-performing influencer ads at scale. No creator searching, no hassle—just results. Book a free demo call by March 7th & claim your $200 bonus!

💬Google’s long-rumored AI assistant, Pixie, is finally launching with the Pixel 10 under the name Pixel Sense. Designed to offer a deeply personalized experience, it will leverage Google services to act as a truly smart on-device assistant.

🔍Microsoft launched Dragon Copilot, an AI assistant for healthcare that transcribes clinical visits, automates documentation, and provides medical insights. Powered by Nuance’s voice tech, it aims to reduce clinician burnout, with 93% of patients reporting better experiences and improved efficiency.

📄 Meta expands anti-scam facial recognition to the UK, allowing public figures to opt into protection against impersonation scams. The video selfie verification tool also launches, addressing security concerns. Notably, Meta previously paid $1.4 billion over biometric data lawsuits.

👻Tencent’s AI bot Yuanbao overtakes DeepSeek as China’s top iPhone app, signaling a heated AI race. With three of the top five free apps being AI chatbots, Tencent, ByteDance, and Alibaba are intensifying competition, with Alibaba investing $53 billion in AI.


That’s all for today’s edition see you tomorrow as we track down and get you all that matters in the daily AI Shift!

If you loved this edition let us know how much:

How good and useful was today's edition

Login or Subscribe to participate in polls.

Forward it to your pal to give them a daily dose of the shift so they can 👇

Reply

or to participate.