Elevate Your Investment Strategy
Harness the power of our AI-driven insights and unique geopolitical expertise for optimal AI portfolio management
Signup Today!
H.E. Abhisit Vejjajiva, the 27th Prime Minister of Thailand. See his endorsement over our quantum optimization portfolio investment project.
The Global Macro Investment Strategy (abbreviated as "Macro") was once the preferred investment approach for hedge funds during the 1990s and 2000s. However, with the conclusion of the Cold War era, its popularity waned as it evolved into a more passive, beta-focused strategy, capitalizing on lower interest rates. Nevertheless, the resurgence of global tensions, exemplified by the Ukraine-Russia conflict and escalating superpower rivalries, suggests a potential revival of the Macro strategy.
A critical aspect that bolsters this revival is the recognition of geopolitical risk, which can significantly and enduringly impact investment portfolios. Despite its importance, this type of risk is neither thoroughly researched nor well understood. While risk management systems and methodologies have seen dramatic evolution since the 1970s, there has been negligible advancement in how market practitioners analyze and manage geopolitical risks. Currently, there are no geopolitical equivalents to commonly used market risk models like value-at-risk (VaR), nor are there analytical frameworks that link geopolitical "risk factors" to expected returns.
Effective engagement with Macro requires deep understanding of geopolitics. Our Meta-Geopolitical Framework offers insightful analysis of the dynamics among global political actors. At the micro level, we apply the concept of "positive governance" to conduct thorough analyses. This approach acknowledges the significant gap in current risk management regarding geopolitical factors. Integrating these insights with advanced AI, we aim to optimize portfolio management, pioneering in an area where traditional models fall short.
Our Journey
The inception of "Midas Project," an AI-driven investment venture, was at the Womanium Quantum Hackathon 2022. Originally focused on using quantum computing for investment optimization, we created a QAOA algorithm-based prototype. However, the unveiling of ChatGPT and GPT-3.5 in November 2022 pivoted our direction towards Generative AI. We're now harnessing this tech to analyze social media and news, feeding it into our Geopolitical module and scenario planning unit. Our goal? To minimize errors and boost accuracy using a vector database, Retrieval-Augmented Generation (RAG), prompt engineering, and our innovative multi-headed, "AI Blue" Project. Our journey includes notable strides in hackathons, securing six finalist spots in the mid-2023 Google Vertex AI Hackathon. Concurrently, we're delving into positive governance research to enhance our meta-geopolitics framework. The Midas Project is in the final stage of our beta development phase.
Adding to our repertoire, we proudly announce our another endeavor, GRASUP (Large Language Models for Geopolitical Risk Analysis via Supercomputing), a geopolitical fine-tuning LLMs. This project has been greenlit for an Open Beta Test under the LANTA programme. LANTA is synonymous with the Cray EX Supercomputer, a technological marvel situated in Thailand. Ranked 79th globally, it is the fastest supercomputer in the ASEAN region and among the top 20 in Asia. For an in-depth look into this groundbreaking project, explore https://thaisc.io.
Our crowning achievement is the Geopolitics GPT, a specialized configuration on the OpenAI's GPT Store, leading the field in the "geopolitics" category with 1K+ sessions usage and growing. We've also integrated the Geopolitics Chatbot at Hugging Face, featuring Geopolitics mxGPT, powered by NousResearch's Nous-Hermes-2-Mixtral-8x7B-DPO. Additionally, we explore the impact of AI personas on chatbot behavior in our experiment to emulate the character Frieren from "Frieren: Beyond Journey's End" manga and anime on Character.AI.
Our article, "Generative AI's Role in Social Science: Research Assistance and Analysis," was published in the PAAT Journal around December 28, 2023, Written in Thai, we aim to release an English version, adhering to the stringent standards of AI or Computational Social Science (CSS) journals. The article underscores how DARPA projects like the Ground Truth Project and NGS2 have propelled computational simulations in social science, highlighting their significant role in the advancement of computational social science and the emergence of Generative AI recently.
The Comprehensive AI Platform
The core AI engine is expected to be a highly intelligent,
all-encompassing Artificial General Super-intelligence (AGI) system that leverages Autonomous Generative AI architecture. The system provides solutions to a wide range of complex tasks while allowing for rapid decision making and autonomous actions.
We have started with the trend radar, PulsarWave. Even in its early stages, it has the potential to generate several business opportunities. For example, it could be used to build AI news anchors and news announcements that relate to the core evolvement of geopolitics around the world. This would be valuable for policymakers to consider for risk assessment, as well as for investors to adjust their wealth management strategies accordingly.
Please download our whitepaper here.
Please see our content talking about adapting AI-Driven, Agile Organizational Tranformation Concept at our affiliated website at http://www.geopolitics.asia, as well as an example of trend radar at http://www.geopolitics.io.
Our Successful Hackathon Project
Our PulsarWave Project has been 6 finalists at Google Vertex AI Hackathon with Lablab.AI
Our users love our product (both on PulsarWave and AI Blue Hackathons)
Testimonials
Don't just take our word for it, read from our extensive list of case studies and user testimonials.
Great idea and implementation! The presentation is decent, butI feel it could have been better. Some points about specific vertex ai tech that was used should have been included in the presentation, and a person should have briefly explained the project before the demo.
Chinmay Jog - Machine Learning Engineer
Very thoughtful prompting throughout the application. Was this application created within the timeframe? I see more OpenAI API then Google Vertex.
Ervin Moore - PhD Computer Science Student
AI Blue is a cool use of new LLM techniques and routing to respond to complicated, nuanced questions not typically answerable with commercial LLMs like chatgpt. Areas of improvement: - It was unclear from a user perspective what role the AI Blue gradient has in influencing the answer. - It would be great to see evaluations used to compare the different approaches and really display the value of your innovative approach.
Josh Reini - CEvents attended
idea is smart enough to work more on it. more efforts are needed to explain what blue gradient means. what could be benefits of this project and how it will affect future of ai models. working on your demo, no correct answer to my questions and many hallucinations.
Walaa Nasr Elghitany - Data scientist and doctor
Our Location
Contact us at our new address:
Siam Intelligence Unit Co., Ltd.
23/F, M. Thai Tower, All Seasons Place,
87 Witthayu Rd, Khwaeng Lumphini,
Pathum Wan, Bangkok.
10330
DBD's corporate ID/Tax ID: 01055-52007-97-1
Registered Consultant, Corporate Type, Recorded on Ministry of Finance, No. 6389
Email: info@siamintelligenceunit.com
Phone: +662-627-9064 Fax: +662-022-0462