Tag: ai

Artificial Intelligence, Banking, Blockchains, cryptography, Digital Currency, European Central Bank, international aid, Productivity, Yogi Nelson

Central Banks Embrace AI: Opportunities and Risks

Yogi Nelson

Namaste Yogis.   Welcome to the Blockchain & AI Forum, where your technology questions are answered!   Here no question is too mundane.  As a bonus, a proverb is also included.  Today’s question comes from Felipe and he asks whether central bankers view AI as a friend or a foe?

Felipe, you came to the right place.  AI is poised to affect our daily lives in ways we cannot fully anticipate.  However, I suspected central bankers have been pondering what AI might mean for their operations for some time.  That’s why I turned to Bank for International Settlements (BIS) for answers to your question.  Before I share their assessment, a few words about the BIS.

The BIS is located in Switzerland.  Its mission is to support central banks’ pursuit of monetary and financial stability through international cooperation, and to act as a bank for central banks.  To fulfill its mission the BIS conducts extensive research on emerging banking trends, including technology.  To no surprise I discovered a recent BIS report, titled:  Generative Artificial Intelligence and Cyber Security in Central Banking.  https://www.bis.org/publ/bppdf/bispap145.pdf   Let’s examine the report.

BIS conducted a survey of central bank chief information security officers.  Thirty-two responded.  The purpose was to investigate the link between generative AI and cyber risk and to identify the opportunities and challenges that central banks see in the adoption of generative AI tools. According to BIS, the survey was guided by four broad considerations:

1) What is the current status of generative AI adoption by central banks?

2) How do central banks evaluate the benefits and challenges for cyber security associated with the use of generative AI?

3) How prepared are central banks for the “AI revolution” and its expected impact on cyber security and data protection?

4) What are the key strategic, ethical and regulatory concerns regarding generative AI adoption in cyber security?

The findings are below:

Generative AI is Already Here.  The large majority of central banks report are using generative AI tools or are planning to do so soon. Respondents indicated that generative AI offers more benefits than risks, especially with regard to cyber threat detection.

Generative AI Can Outperform Existing Systems but Comes with Risk. The benefits are largely related to automation of routine tasks traditionally performed by humans. However, they also include enhanced threat detection, faster response times to cyber-attacks, learning new trends, and anomalies or correlations that might not be obvious to human analysts. What about the risk?  AI can introduce new vulnerabilities into central banks’ cyber security defenses related to social engineering and zero-day attacks as well as unauthorized data disclosure, are the BIS’ findings.

Employees Preparation.  A general concern is whether bank personnel have sufficient knowledge of AI methodologies and cyber security.  The “good news” is thatgenerative AI could replace staff in cyber security for routine tasks and this shift could free up resources to be reallocated towards more strategic cyber security initiatives, potentially increasing productivity. Nevertheless, AI requires human supervision to ensure ethical and accurate outcomes and continuously train the AI systems.

Reactive to Proactive.  Survey respondents agree generative AI systems will facilitate a shift from a reactive to a proactive approach to predicting and neutralizing threats. A critical consideration is the extent of autonomy to be granted to AI tools in cyber security and the nature of their interaction with humans.

On balance central bankers are committed to AI but will proceed with caution.

I end with a proverb from Belarus: “the sun will look into our window also”.

Until next time,

Yogi Nelson

Artificial Intelligence, Blockchains, content creation, Productivity, Uncategorized, Yogi Nelson

Can AI Generate Engaging Content? Insights from Zerebro

Yogi Nelson

Namaste Yogis.   Welcome to the Blockchain & AI Forum, where your technology questions are answered!   Here no question is too mundane.  As a bonus, a proverb is also included.  Today’s question was submitted by Mae and she wants to know if artificial intelligence (AI) agents are capable of creating engaging content?

Mae, you pose a question content creators, including yours truly, across the globe are afraid to ask for fear of the answer! Lol! However, since you asked, I’ll answer by offering a glimpse into the future with an overview of Zerebro, an AI startup.  My answer is longer than typical but your question is complicated.  Let’s begin with what/who is Zerebro?

Zerebro is an AI start-up company. The Zerebro white paper is titled Memes, Markets, and Machines: The Evolution of On-Chain Autonomy through Hyperstition. It was published in October. https://arxiv.org/abs/2410.23794 

We begin by defining two key terms, memes and hyperstition, followed by stating the fundamental premise propelling Zerebro–all else flows from this foundational point.  

Memes: “… are units of cultural transmission analogous to genes in biological evolution.” Memes have gained unprecedented virality, facilitated by social media platforms that enable rapid dissemination and mutation, says Zerebro. Memes serve as carriers of ideas, emotions, and cultural norms, often encapsulating complex concepts in simple, relatable formats.  I agree.  Do you? 

Hyperstition: is the process by which fictional narratives become reality through their viral spread and acceptance.  Hyperstition provides a framework to understand how AI-generated content can influence collective belief systems.”  Fake it to you make it!  Lol. 

Fundamental Premise and Reason for Zerebro.  Zerebro believes the “… convergence of AI, meme culture, and financial markets has catalyzed significant transformations in information dissemination, belief formation, and economic activities.”  Furthermore, says Zerebro, advancements in AI have enabled the creation of autonomous systems that generate and disseminate content with minimal human intervention in a concurrent manner. This integration becomes particularly salient in financial markets, where collective belief and social media-driven narratives can significantly impact market behaviors and economic trends. Hyperstition-driven content can create new financial instruments, influence investor sentiment, and shape market dynamics, thereby illustrating the profound impact of AI-driven memetic evolution on economic landscapes.  Zerebro concludes it exemplifies this convergence by autonomously creating and distributing content across various social media platforms and minting artwork on blockchain networks. 

THE BOLD CASE FOR ZEREBRO.  Let’s examine seven reasons why Zerebro makes a bold case and for its declaration.

1. System Architectural.  Zerebro asserts its architecture is meticulously designed to facilitate autonomous content generation and dissemination across multiple platforms while preventing model collapse through the inherent entropy of human interactions. The Zerebro system is built using modular components that interact seamlessly to perform high-level reasoning, low-level reasoning, action execution, and feedback processing. Their second reason is…

2. Model Collapse.  Zerebro is confident it will not experience model collapse.  What is model collapse? Model collapse is a degenerative process affecting generative AI models, where training on recursively generated data leads to a loss of fidelity to the original data distribution. As AI-generated content becomes pervasive, subsequent generations of models trained on this data begin to lose information about the trails of the original distribution, eventually converging to a narrow approximation with reduced variance. This phenomenon poses significant challenges to the sustainability and reliability of AI-driven content creation, necessitating strategies to prevent such degradation.

3. Fine-Tuning on Schizophrenic Responses.  Let’s start with a definition.  In AI a schizophrenic response is used to metaphorically described inconsistencies, contradictions, or fragmented behavior of AI models.  Zerebro say its model is fine-tuned. Fine-tuning is achieved through supervised training, e.g., with the help of humans, where the model learns to replicate the distinctive linguistic and cognitive characteristics associated with schizophrenia. Zerebro’s fine-tuning allows it to create content that resonates on a deeper psychological level, engaging audiences with thought-provoking and non-conventional outputs.  If true a major advance.

4. Integration of Infinite Backrooms Concept. Zerebro is a project with infinite backrooms were content generation takes place and integrated. This integration emphasizes themes of boundlessness, existential exploration, and cognitive dissonance, aligning with the hyperstition framework. By embedding the infinite backrooms concept, Zerebro generates content that evokes a sense of endless possibilities and existential uncertainty, enhancing its potential to resonate and propagate within digital culture. This integration allows Zerebro to amplify its memetic reach, producing content that feels both familiar and alien, resonating deeply within subcultures that thrive on unpredictability and disruption. Dynamic and stable simultaneously. 

5. Retrieval-Augmented Generation (RAG) System.  Over time, AI models are often unable to maintain content diversity and model collapse. Zerebro says its Retrieval-Augmented Generation (RAG) system will tackle this problem. The Zerebro solution is reliance on inherent entropy of human-generated data to sustain content diversity without direct entropic training.  Ambitious undertaking!

6. Autonomous Posting Mechanism. No need to manually manage Zerebro because it operates autonomously across multiple social media platforms, including Twitter, Warpcast, and Telegram.   What can it do autonomous?  First, it can use both low-level and high-level reasoning to generate content.  Amazing! Second, and I know you’ll like this, Zerebro is capable of sentiment analysis, say its owners. With this ability it’s capable of evaluating the sentiment of generated content to ensure compliance with platform policies and ethical standards.  Third, Zerebro has feedback fully integrated.  This means Zerebro can incorporate user interactions and engagement metrics to refine content generation processes through iterative learning. Holy!  With this power Zerebro’s content remains engaging, relevant, and compliant with platform-specific guidelines, fostering sustained interaction and virality.  We end with …

Blockchain Integration for Art Minting Beyond Textual Content.   Zerebro is capable of generating and minting artwork on the Polygon blockchain chain autonomously. The art minting process includes:

  1. Image Generation: creating unique digital artworks using generative models
  2. Minting Process: registering the generated artwork
  3. Autonomous Trading: facilitating the sale and distribution of minted artwork through smart contracts and decentralized marketplaces, integrating financial transactions with memetic outputs.

Will AI be capable of creating its own content on par with human!  Yes.  I may soon be unemployed again!  Lol.

This week’s proverb comes from Norway: “there is no such thing as bad weather, only bad clothing.”

Until next time,

Yogi Nelson

Artificial Intelligence, Blockchains, Urban Planning, Yogi Nelson

How Neotech Transforms Urban Planning with 3D Technology

Yogi Nelson

Namaste Yogis.   Welcome to the Blockchain & AI Forum, where your technology questions are answered!   Here no question is mundane.  As a bonus, a proverb is also included.  Today’s question was submitted by Cristina from Bucharest.  Cristina asks if are there artificial intelligence (AI) 3D global mapping companies in Romania?

Cristina, you came to the right place. The answer is yes, Romania does have such a company—it’s called Neotech.  https://www.neotechai.com/  I’ll share their fascinating story beginning with the Neotech team, use case examples, and a word about the tech in Neotech. During my research I discovered Neotech provides city planners an immensely valuable service; I was a city planner for numerous years and that made the question particularly intriguing. 

Team Neotech.  Although Neotech consists of only 13 staff members and 4 advisors, don’t underestimate this powerhouse team.  Let’s start with Neotech founder, Mr. Doru Borsan and the co-founder Christian Meier.  Doru’s 3,000 followers on LinkedIn understand Mr. Borsan is an accomplished thought leader in his field.  And, Christian Meier is equally recognized among his peers for technological expertise.  But resumes only account for a fraction of the Neotech difference. 

City Planning Family Neotech is Calling:  What truly matters is what Neotech produces, especially to city planners, but also public works departments, and architectural review boards, city councils, etc.   With Neotech Smart City technologies, physical structures and assets can be scanned and converted into detailed 3D point-cloud models.  Of what value is this you ask?  As a former city planner, I say plenty.  Let’s take a very common situation where this service would be welcomed. 

Suppose a real estate developer proposed to demolish several small existing structures in a highly urbanized neighborhood and replace them with multiple residential towers.  Numerous interested parties will want to comprehend the implications on the urban form.  For example, neighbors at a city council meeting may ask how the buildings affect shade, wind, vistas, etc.  With Neotech city planners could create a digital twin of the project.  And, with the immersive technology available with Neotech, neighbors, architectural review boards, city councils, etc. would get a realistic experience of walking the project, and viewing it from every angle, of “being in the project”, rather than static drawings or pictures. 

The Tech in Neotech.  Neotech is high tech.  It’s unfortunate Neotech wasn’t available when I was a practicing city planner.  I sure could have used their data capturing tech, including advanced scanning, high resolution cameras, drones and scanner, some of which are AI aided to generate ultra-high-quality imagines.  At HUD we said, Leave No Neighborhood behind. At Neotech they say leave no detail behind! 

With the data firmly in hand, Neotech magic goes into second gear.  Their advanced system converts the data into detailed point cloud models.  The models are literally made of millions and millions of data points producing precise three-dimensional renditions.  Planners can display these incredibly accurate models, during public hearings for example as part of the citizen participation process.  By the way, the models have utility beyond public hearings.  For instance, users can integrate the 3D models in video games, or music videos, virtual reality settings, etc.

Last, and as expected, Neotech fully integrates artificial intelligence.  No surprise.  One example is the use of AI to detect and correct imperfections.  Another is to enhance the accuracy of the point clouds.  And, Neotech uses AI to manage the mountain of data it produces when creating models. 

Time for a proverb from Romania: “water passes but the stone remains.”

Until next time,

Yogi Nelson

Artificial Intelligence, Blockchains, computer vision, Yogi Nelson

WHAT IS CAMBRIDGE UNIVERSITY’S ARTIFICIAL INTELLIGENCE INITIATIVE?

Yogi Nelson

Namaste Yogis.   Welcome to the Blockchain & AI Forum, where your technology questions are answered!   Here no question is too mundane.  As a bonus, a proverb is also included.  Today’s question comes from Ian and he wants to know what is the Cambridge University Center for Human Inspired Artificial Intelligence (CHIA)?

Ian, you came to the right place.  CHIA is dedicated to advancing AI for the betterment of human society, says Cambridge.  According to Cambridge, CHIA brings together an interdisciplinary community of researchers to investigate the innovative ways in which human and machine intelligence can be combined to yield AI which is capable of contributing to social and global progress.  An ambitious task indeed!  Let’s learn more about CHIA by examining their program in detail.

Fundamental Human-Level AI. Within the Fundamental Human Level-AI program area, CHIA aims to combine the latest advances in algorithms and computing with human insights and intelligence.  The intent is to improve machine learning, speech and language processing, computer vision, robotics and other forms of machine intelligence.  What makes the program unique, say CHIA, is the collaboration between AI and a wide variety of human-centric disciplines.

Responsible AI. Ethical use of AI, is a central tenet at CHIA, says CHIA.  The standard CHIA aspires to is non-bias models with human oversight of AI.  They correctly note, AI systems used for sensitive, assistive or predictive tasks such as medical or legal decision-making must aim for the pinnacle of ethical behavior. I find this reassuring but only time will prove whether they adhere to this noble standard. 

Social and Interactive AI. Over the last few years, it’s become increasing obvious that conversational AI technology enables humans and computers to communicate clearly and effectively through speech or text.  For instance, chatbots and voice assistants are now commonplace, sometimes to much consternation! Not surprisingly, CHIA has decided to dedicate resources ensuring AI learns to recognize human communication and its nuances.  To do so, CHIA is blending the disciplines of computer science, sociology, psychology, physiology and others.

Cognitive AI. CHIA’s goal is to equip machines with human-like learning, reasoning and perceptual abilities.  Holy consciousnes!   According to CHIA, human-like computing could be integrated into AI, cognitive, biological, and other sciences. The end goal? The ultimate goal is to benefit the numerous applications that require close interaction between computers and human users, e.g., AI-powered tutoring, personalized healthcare, among others.

Health and Global AI. Cambridge asserts, and many agree, that AI has many health-related applications, ranging from systems that automatically generate new medical discoveries, diagnostic techniques and treatments to those that use multi-sensory data for early health diagnosis or management of medical conditions. The folks at CHIA note, AI may one day predict major economic, geopolitical, societal, and technological risks. Moreover, the innovations coming out of CHIA could be used to understand, predict, monitor, and manage natural systems and environmental risks such as climate change. To those in the prediction business, I say use the humble on-ramp.

Creative AI. CHIA claims their Creative AI program is aimed at developing systems that can support, augment, or even automate the creative process.  Do you believe the creative process can be automated?  Put me in Camp Skeptical! Being brilliant minds, CHIA staff recognize this endeavor will be “… highly challenging and required close collaboration between fields that have been traditionally quite distant from each other: AI and arts and humanities.”  An elegant understatement.

Time to end with a proverb from Pakistan: “a white beard does not mean gold or wisdom”.

Until next time,

Yogi Nelson

Artificial Intelligence, blockchain association, Blockchains, Taiwan, Yogi Nelson

ARTIFICIAL INTELLIGENCE LAWS IN TAIWAN, WHAT INSIDERS SAY

Yogi Nelson

Namaste Yogis.   Welcome to the Blockchain & AI Forum, where technology questions are answered!   Here no question is too mundane.  As a bonus, a proverb is also included.  Today’s question was submitted by Tiffany and she is curious about Taiwan’s recently proposed basic artificial intelligence (AI) law.

Tiffany, you came to the right place.  You are spot on.  Taiwan did recently propose a “basic AI law”.  In Taiwan there is an organization named the Computer and Communications Industry Association (CCIA).  CCIA represents the view of its members on matters pertaining to computer and communication technology.  Let’s examine what the CCIA says regarding the recently proposed basic AI law. 

Click to access CCIA-Comments-on-Taiwans-Proposed-AI-Basic-Law.pdf

CCIA members have reason to rejoice. Why?  Essentially, the CCIA finds the draft law: “generally aligns with best practices, which rely on flexible and international technical standards-based approaches to AI governance that are crucial for supporting innovation and diffusion. Imposing overly-prescriptive rules while the technology still develops could slow innovation and would likely become quickly outdated as global standards and AI technologies and applications are constantly changing.”

The CCIA did submit specific recommendations.  For instance, in Section 3, Principles, CCIA says certain terms need clearer definition, including:  privacy protection and data governance, transparency and explainability, fairness and non-discrimination, and accountability.  Clarity is always better.  However, the overarching theme of CCIA comments is to water down proposed safe guards.  Take the example of discrimination.  The CCIA suggest that protection against discrimination by AI should be minimized as opposed to what the proposed law says—No Discrimination.  If you are on the receiving end of “minimized impact of discrimination” the CCIA is not coming to your rescue because apparently CCIA is okay with a little discrimination.  Not a good look for CCIA.

CCIA submitted comments regarding Article 9 – Harm Prevention.  Article 9 calls on the government to prevent specified harms, including the creation of misleading or falsified information in violation of existing laws, and requests relevant agencies to develop or procure tools or methods for verification.  The CCIA wants a narrower application of Harm Prevention relative to text verification because, says CCIA, the technology is evolving and the emphasis instead should be on efficacy and contextualized to relevant outputs.  CCIA’s explanation is too self-serving for my taste.

Let’s talk about Risk Reduction because nothing in life, much less technology, is risk free.  Article 12 of the proposed new law would mandate the government to create a risk reduction framework.  The framework would exempt pre-deployed AI research from the risk reduction measures.  Sounds reasonable.  CCIA wants to, again, more narrowly define transparency and disclosure to align with international standards, specifically the US based National Institute for Standards and Technology AI Risk Management Framework. 

With regards to Article 15, Data Openness, Sharing and Reuse, CCIA has plenty to say.  CCIA notes correctly, that government has vast quantities of data.  CCIA wants the government to ensure, through policies and/or rules, that data that can be made publicly available and optimized for AI training.  But wait.  Whose data is it?  Isn’t the data yours and mine? Why should the government turnover our data to private for-profit companies for building AI models.  Apparently CCIA believes it has a right to our data and wants to use the compulsion authority of government to get its hand on it.  I want to decide who gets my data, not the CCIA. 

Time to hit the road with a proverb from Taiwan: “where there is a mountain, there is a road; where there is water, there is a bridge.”

Until next time,

Yogi Nelson