A Daily Chronicle of AI Innovations on March 18th 2025
Baidu introduced new, competitively priced AI models, challenging existing market leaders in both performance and cost. A court upheld OpenAI's right to continue its AI development, dismissing Elon Musk's concerns. Harvard researchers unveiled an AI agent for personalised medicine, demonstrating progress in healthcare innovation. Google's new AI model showed impressive watermark removal capabilities, raising questions about copyright protection. Furthermore, AI is increasingly being integrated into hospital care, facing resistance from human nurses, and experts predict AI will soon surpass human coders.
🇨🇳Baidu Launches Ultra-Cheap AI Models
Chinese tech giant Baidu has unveiled two new AI models: ERNIE 4.5 and ERNIE X1. ERNIE X1 is positioned as a cost-effective competitor to DeepSeek’s R1 model, offering comparable performance at half the price. Both models are now available for free to individual users of Baidu’s chatbot services.
ERNIE 4.5 stands out with enhanced “EQ” and language skills as well as improved hallucination prevention, logical reasoning, and coding abilities.
According to Baidu, ERNIE 4.5 outperforms GPT-4o across multiple benchmarks while costing just 1% of its price — approximately $0.55 and $2.20 per million input and output tokens.
ERNIE X1, Baidu’s first reasoning model, matches the capabilities of top Chinese competitor DeepSeek's R1 at half the price.
Like DeepSeek’s R1, ERNIE X1 uses a step-by-step "thinking" approach, excelling in complex calculations and tasks like document understanding.
What this means: Baidu’s aggressive pricing strategy intensifies the AI market competition, potentially prompting other companies to reevaluate their pricing and accessibility to maintain market share.
⚖️ Judge Rejects Musk’s Bid to Block OpenAI’s Evolution
A recent court ruling has denied Elon Musk’s attempt to halt OpenAI’s development trajectory. Musk expressed concerns over the rapid advancement of AI and potential risks associated with it. However, the court found no legal basis to impede OpenAI’s progress.
The judge said Musk didn’t have "the high burden” for a preliminary injunction but allowed other aspects of the lawsuit with an expedited trial this fall.
Internal emails cited by OpenAI allegedly show Musk once wanted to merge OpenAI into Tesla as a for-profit entity—contradicting his current legal position.
Musk sued OpenAI and its CEO, Sam Altman, last year over straying from their original mission of developing AI for the good of humanity, not corporate profits.
OpenAI and Altman have denied the accusations, saying the non-profit is not going—and any restructuring of for-profit subsidiaries will only better support it.
What this means: The decision underscores the judiciary’s current stance on AI development, balancing innovation with ethical considerations, and highlights the challenges in regulating rapidly evolving technologies.
🧬 Harvard Team Creates AI Agent for Personalized Medicine
Researchers from Harvard and MIT have introduced TxAgent, an AI agent designed to provide personalized treatment recommendations. TxAgent leverages multi-step reasoning and real-time biomedical knowledge retrieval across a toolbox of 211 tools to analyze drug interactions, contraindications, and patient-specific treatment strategies.
TxAgent utilizes 211 specialized tools to analyze drug interactions and contraindications and suggest patient-specific treatments in real time.
It evaluates drugs at molecular, pharmacokinetic, and clinical levels, identifying risks based on comorbidities, ongoing medications, age, and genetic factors.
The system synthesizes evidence from biomedical sources, iteratively reasoning and refining recommendations through structured function calls.
TxAgent's toolkit is available as "ToolUniverse" featuring trusted data sources like openFDA and Open Targets to ensure validated medical insights.
What this means: TxAgent represents a significant advancement in precision medicine, offering clinicians AI-driven insights to tailor treatments to individual patient needs, potentially improving outcomes and reducing adverse effects.
💥 Baidu Launches ERNIE 4.5: A GPT-4.5 Rival at 1% the Cost
Chinese tech giant Baidu has unveiled ERNIE 4.5, a multimodal AI model designed to compete with OpenAI’s GPT-4.5. Baidu claims that ERNIE 4.5 not only matches but surpasses GPT-4.5 in performance metrics, all while operating at just 1% of the cost. This aggressive pricing strategy positions ERNIE 4.5 as a highly cost-effective solution for businesses and developers seeking advanced AI capabilities without the hefty price tag.
Baidu has introduced two new foundation models, ERNIE 4.5 and ERNIE X1, claiming ERNIE 4.5 is significantly cheaper than OpenAI's GPT-4.5, costing just 1% of its price.
ERNIE X1 is designed for advanced reasoning and understanding tasks, excelling in Chinese Q&A, logical reasoning, and complex calculations, while being priced at half the cost of DeepSeek R1.
The models are accessible through ERNIE Bot's website, with plans for integration into Baidu's services and availability via APIs on Baidu AI Cloud's platform for developers.
What this means: Baidu’s introduction of ERNIE 4.5 could disrupt the AI market by making high-quality AI services more accessible and affordable. This move may compel other AI providers to reassess their pricing models to remain competitive.
🖼️ Google’s New AI Excels at Watermark Removal
Google has developed an advanced AI model capable of removing watermarks from images with remarkable precision. While this showcases the potential of AI in image processing, it also raises ethical concerns regarding copyright infringement and unauthorized use of protected content.
Social media users have highlighted that Google's new Gemini AI model can erase watermarks from images, including those from well-known stock media companies like Getty Images.
The recently expanded Gemini 2.0 Flash model can generate and edit images, easily removing watermarks while filling in any resulting gaps, showcasing its advanced capabilities.
Though labeled as experimental and not yet for production, the model's ability to remove watermarks raises concerns about potential copyright violations, as it lacks usage restrictions unlike other AI models.
What this means: The ability of AI to remove watermarks challenges existing methods of protecting digital content. This development may prompt content creators and platforms to seek more robust watermarking techniques or alternative methods to safeguard intellectual property.
🏥 AI Nurses Are Reshaping Hospital Care, but Human Nurses Are Pushing Back
Hospitals are increasingly integrating AI-powered nurses to assist with patient monitoring, medication management, and administrative tasks. These AI systems can provide real-time updates on patient vitals and automate routine procedures. However, human nurses are resisting the rapid deployment of AI, arguing that it depersonalizes patient care and could lead to staff reductions.
What this means: AI-driven healthcare innovations could streamline hospital operations, but the pushback from human nurses underscores concerns about job security, ethical implications, and patient trust. Future AI-driven healthcare models may need to strike a balance between efficiency and human-centered care.
🖼️ People Are Using Google’s New AI Model to Remove Watermarks from Images
Google’s latest AI model, initially designed for image enhancement, has been widely adopted for watermark removal, sparking concerns among photographers, artists, and content creators. The AI can analyze and reconstruct altered pixels, effectively eliminating digital watermarks meant to protect copyrighted materials.
What this means: This raises major ethical and legal questions about AI’s impact on content protection. As AI capabilities improve, watermarking techniques may need to evolve to remain effective, and new regulations could emerge to prevent AI misuse in intellectual property infringement.
🔬 Discovering Organic Reactions with a Machine-Learning-Powered Deciphering of Tera-Scale Mass Spectrometry Data
Scientists have developed an AI-driven system capable of decoding tera-scale mass spectrometry data to identify new organic chemical reactions. This breakthrough could significantly accelerate drug discovery, material science innovations, and synthetic chemistry research.
What this means: AI is transforming scientific discovery by making complex data processing faster and more efficient. This new approach could revolutionize drug design, energy solutions, and industrial chemistry, reducing research time from years to months or even weeks.
💻 OpenAI CPO Kevin Weil Predicts AI Will Surpass Human Coders by Year-End
Kevin Weil, OpenAI’s Chief Product Officer, has predicted that by the end of 2025, AI models will surpass human coders in speed, accuracy, and innovation. Weil argues that AI coding assistants have already begun outpacing human developers in areas such as software optimization, debugging, and automation.
What this means: If AI achieves this milestone, it could disrupt software development, forcing engineers to adapt by focusing on high-level architecture, AI oversight, and strategic planning. The shift may also change hiring practices, with companies prioritizing AI-literate engineers over traditional coders.
🧬 Bespoke Cancer Vaccines Are Suddenly Looking Extremely Promising
Researchers are making major breakthroughs in the development of mRNA-based cancer vaccines, which could revolutionize personalized cancer treatment. Unlike traditional vaccines, these custom-made therapies train the immune system to recognize and attack specific cancer cells unique to an individual patient.
Recent clinical trials have demonstrated significant tumor shrinkage and increased survival rates in patients receiving personalized mRNA cancer vaccines. Pharmaceutical giants Moderna and BioNTech are leading the charge, with new vaccines targeting a range of cancers, including melanoma, lung, and pancreatic cancer.
What this means: AI-driven drug discovery and advancements in mRNA technology could make personalized cancer vaccines a standard treatment in the near future. If successful, this approach could drastically improve survival rates and eliminate the need for aggressive chemotherapy in some patients. The rapid progress also raises hopes for applying similar methods to other diseases beyond cancer.
What Else Happened in AI on March 18th 2025?
Google’s new Gemini 2.0 Flash model is reportedly being used by people to remove watermarks from images, including those from platforms like Getty.
Humanoid robot maker Figure announced the launch of BotQ, a new manufacturing facility capable of producing 12,000 humanoids a year—with plans to scale up to 100,000.
Patronus AI launched the industry’s first multimodal LLM-as-a-judge, a tool designed to help developers detect and mitigate reliability issues in multimodal AI apps.
Pika Labs released 16 new effects for its AI video platform, enabling users to morph images into different character videos.
Sesame, which recently went viral for its realistic AI voice tech demo, open-sourced its Conversation Speech Model (CSM-1B) for text-to-speech tasks.
Vogent AI launched voice agents that design and improve themselves, learning from real failure cases and requiring no prompt engineering.
Y Combinator CEO reported that nearly a quarter of their current startups now have 95% of their code written by AI, enabling small teams to reach $10M in revenue with dramatically fewer engineers.
OpenAI's CPO Kevin Weil just predicted 2025 is the year AI will permanently surpass humans at programming, calling it a "democratizing effect" that will allow anyone to create software.
🚀Advertise on AI Unraveled: Reach Thousands of AI Enthusiasts Daily!
AI Unraveled is your go-to podcast for the latest AI news, trends, and insights, with 500+ daily downloads and a rapidly growing audience of tech leaders, AI professionals, and enthusiasts. If you have a product, service, or brand that aligns with the future of AI, this is your chance to get in front of a highly engaged and knowledgeable audience. Secure your ad spot today and let us feature your offering in an episode!
🎙️ Book your ad spot now: https://buy.stripe.com/fZe3co9ll1VwfbabIO
🙏 Support the AI Unraveled Podcast and Channel:
Consider buying me a coffee to say thank you for the free tech content on my YouTube channel (@enoumen) and the AI Unraveled podcast. https://buy.stripe.com/3csaEQ1ST9nYgfe4gk
Share this post