⚠️ Finance - Maharlika Investment Fund's Investment in Celsius Resources
This article analyzes the Maharlika Investment Fund's $76.4M USD investment in Celsius Resources (ASX: CLA), highlighting potential red flags based on the company's financial performance.
Key Points:
• Zero revenue reported.
• Significant net loss of -$8.43M AUD.
• Negative cash flow of -$5.84M AUD.
• Public funds are at risk due to this investment.
Image
Image
Image
Image
🔗 Resources:
• Mong Ramos Jr. ↗ - Author of the tweet
• Allan Jacob ↗ - Source of the financial data
🤖 Conversational AI - Advanced Natural Language Dialogue
This article discusses a new conversational AI system demonstrating a significant leap in naturalness compared to existing solutions like Advanced Voice Mode or Grok 3. A demo version is linked for testing.
Key Points:
• Unprecedented naturalness in conversation.
• Maintains coherent conversation over extended periods.
• Significantly surpasses the naturalness of existing systems.
Image
🔗 Resources:
• Florian Gallwitz ↗ - Source of the demo
💡 Peer Review - Issues with Meta-reviews and Reviewer Comments
This article describes a frustrating experience with peer review where the meta-review and final reviewer comments were factually incorrect, leading to paper rejection despite the presence of real-world data.
Key Points:
• Meta-review and final reviewer comments contained factual inaccuracies.
• Paper rejected despite inclusion of two real-world datasets.
• Novelty, a subjective criterion, was cited as a reason for rejection.
🔗 Resources:
• Kwangmoo Yi ↗ - Author of the tweet
🤖 Robotics - Mast3r SLAM Integration with Rerun
This article discusses the integration of the Mast3r SLAM library with the Rerun visualization tool for improved usability and time-series data handling.
Key Points:
• Current implementation uses ImGui and GL with custom shaders.
• Rerun aims to improve ease of use.
• Time-series visualization capabilities are a key benefit of using Rerun.
🔗 Resources:
• Pablo Velagomez ↗ - Author of the tweet
• rerun ↗ - Visualization tool
⚠️ Peer Review - Unfair Review Process at CVPR
This article describes an unfair peer review process at CVPR, where an irresponsible reviewer's actions influenced other reviewers' ratings.
Key Points:
• An irresponsible reviewer was reported.
• Other reviewers changed ratings based on the irresponsible reviewer's comments.
• The resulting review process deemed unfair.
🔗 Resources:
• Faegheh Sardari ↗ - Author of the tweet
• CVPR ↗ - Conference
🚀 AI - Active Agent for Rails AI Feature Development
This article discusses Active Agent, a Rails framework designed to simplify the development and maintenance of AI features within Rails applications.
Key Points:
• Streamlines AI feature development in Rails.
• Employs standard Rails conventions for ease of use and maintainability.
• Reduces code complexity.
🔗 Resources:
• TonsOfFun111 ↗ - Author of the tweet
• Inazarova ↗ - Source of information on Active Agent
🤖 Robotics - Robust Manipulation Policies without Extensive Funding
This article highlights the development of robust robot manipulation policies without relying on significant funding or complex techniques like teleoperation, sim2real transfer, generative AI, or motion planning.
Key Points:
• Robust manipulation policies achieved without large datasets or complex methods.
• No teleoperation, sim2real, generative AI, or motion planning used.
• Point Policy is the core technology.
Image
🔗 Resources:
• Lerrel Pinto ↗ - Author of the tweet
• Siddhant Haldar ↗ - Developer of Point Policy
🤖 Robotics - Hardware Limitations in Robotics
This article argues that hardware is not, and will not be, a limiting factor in robotics development.
Key Points:
• Hardware is not a limiting factor in robotics.
• Even with reduced tactile feedback, complex tasks are achievable.
Image
🔗 Resources:
• Pablo Velagomez ↗ - Author of the tweet
• Chris Paxton ↗ - Source of the image
🤖 AI - GPT-4.5 Hallucination Rate Reduction
This article discusses the significant reduction in hallucination rate observed in GPT-4.5, highlighting its implications for model reliability and usability.
Key Points:
• Significant drop in hallucination rate from 61% to 37%.
• Improved reliability and reduced need for fact-checking.
Image
🔗 Resources:
• Giffmana ↗ - Author of the tweet
• Slow Developer ↗ - Source of the hallucination rate data
🚀 Tools - Gemini 2.0 for Content Summarization
This article showcases a new use case for Gemini 2.0: creating Marp slide decks from various content types (PDFs, codebases, etc.) in under a minute.
Key Points:
• Generates Marp slide decks from various input types (PDFs, code repositories).
• Creates beautiful overviews quickly.
• Supports multiple input formats simultaneously.
Image
🔗 Resources:
• Lakshya ↗ - Author of the tweet
• DynamicWebPaige ↗ - Source of the use case
• Code ↗ - Markdown editor
⭐️ Support
If you liked reading this report, please star ⭐️ this repository and follow me on Github ↗, 𝕏 (previously known as Twitter) ↗ to help others discover these resources and regular updates.
