Baidu Unveils Ernie 4.5 & Ernie X1

Google's Gemini 2.0 Flash Raises Copyright Concerns

Author

Jahanzaib Ahmed
March 17, 2025

Welcome AI Fans!
Top story: Chinese search giant Baidu has launched two new AI models — Ernie 4.5 and Ernie X1 — in a move that intensifies competition with Western AI companies.

———————————————————————

In today’s insights:

  • 🔥 Baidu Unveils Ernie 4.5 & Ernie X1

  • ⚡ How to Automate Workflows with CrewAI?

  • 🗣️ Google's Gemini 2.0 Flash Raises Copyright Concerns

  • 🛠️ New Top AI Tools

  • 🔮 More in the AI’s World

🕒 Estimated Reading Time: Under 4 minutes

LATEST IN AI

Image source: Neowin

Overview: Chinese search giant Baidu has launched two new AI models — Ernie 4.5 and Ernie X1 — in a move that intensifies competition with Western AI companies. The release on March 16 positions Baidu as a formidable player in the global AI race, particularly in the reasoning model space.

Essential Details:

  • Ernie X1 is Baidu's first multimodal deep-thinking reasoning model, claimed to match DeepSeek R1's performance at half the price.

  • Ernie 4.5, a multimodal foundation model, reportedly outperforms OpenAI's GPT-4.5 on multiple benchmarks while costing just 1% of GPT-4.5's price.

  • Both models feature multimodal capabilities, processing video, images, audio, and text.

  • Baidu plans to make its Ernie Bot available to the public for free starting April 1, ahead of schedule.

  • Enterprise users can access Ernie 4.5 through Baidu AI Cloud's Qianfan platform, with prices starting at 0.004 RMB per thousand input tokens.

Why It Matters: Baidu's new AI models highlight China's growing influence in the global AI landscape. The significant price advantage over Western competitors could disrupt the market and accelerate the shift toward more affordable AI solutions. This development may pressure Western companies like OpenAI and Google to reconsider their pricing strategies as the competition intensifies.

Overview: CrewAI is a powerful multi-agent automation platform that enables businesses to streamline workflows using AI-powered agents. It allows users to build, deploy, and track AI-driven automations that integrate seamlessly with any LLM (Large Language Model) and cloud platform.

Step-by-step:

  • Sign up and choose a multi-agent automation template.

  • Customize the AI workflow, defining task delegation and tool integrations.

  • Configure deployment options for cloud, self-hosted, or local execution.

  • Activate real-time monitoring to track AI agent performance.

  • Scale automation by refining workflows based on analytics feedback.

Pro tip: Use human-in-the-loop (HITL) where necessary to supervise AI-driven decisions and ensure compliance in sensitive processes.

Image source: Gemini

Overview: Users have discovered that Google's new Gemini 2.0 Flash model can effectively remove watermarks from images, including those from Getty Images and other stock photo providers. This capability, revealed in social media reports on March 16, has sparked concerns about copyright infringement and ethical AI use.

Key Details:

  • Gemini 2.0 Flash not only removes watermarks but attempts to fill in gaps created by the deletion.

  • The model's image generation feature is currently labeled "experimental" and "not for production use."

  • Unlike competitors such as Anthropic's Claude 3.7 Sonnet and OpenAI's GPT-4o, Gemini 2.0 Flash lacks restrictions against watermark removal.

  • Removing watermarks without the original owner's consent is generally considered illegal under U.S. copyright law.

  • The model has limitations, struggling with semi-transparent watermarks and those covering large portions of images.

Why It Matters: Google's apparent oversight in implementing guardrails against watermark removal highlights the ongoing challenges in responsible AI deployment. This issue could lead to increased scrutiny from copyright holders and regulators, potentially forcing Google to implement stronger content policies. The incident also raises broader questions about the balance between powerful AI capabilities and necessary ethical constraints.

AI INSIGHTS

✅ KwiCut – AI for editing talking head videos with voice cloning.
✅ Lazycom-Smart – AI for secure and efficient task management.
✅ Lingolette – AI for language practice through voice and chat.
✅ Made Live – AI for creating illustrated children's books.
✅ Lunroo – AI assistant for social media growth and marketing.

China unveiled Manus, a fully autonomous AI agent capable of executing complex tasks autonomously, developed by Tencent Holdings-backed startup Butterfly Effect in partnership with Alibaba's Qwen.

Larry Page launched Dynatomics, an AI startup focused on next-generation manufacturing, leveraging AI for design automation to create highly optimized product designs.

Elon Musk's AI chatbot, Grok, sparked controversy for using unfiltered and abusive Hindi, mirroring users' aggressive tones and raising concerns about bias and misinformation.

Shield AI raised $240 million at a $5.3 billion valuation to expand its Hivemind Enterprise platform for AI-powered autonomy in robotics and drones.

Andrew Barto and Richard Sutton were awarded the 2024 Turing Award for their groundbreaking contributions to reinforcement learning, shaping AI advancements in robotics and decision-making systems.

Thanks for sticking around…

That’s all for now—catch you next time!

How would you rate our today's Newsletter?

Vote below and help us improve it for you.

Login or Subscribe to participate in polls.

Have any thoughts or questions? Feel free to reach out at community@nurowave.ai – we’re always eager to chat.

P.S.: Do follow me on LinkedIn and enjoy a little treat!

Jahanzaib & The Nurowave Team

Reply

or to participate.