Tag: Blockchain

Artificial Intelligence, Blockchains, cryptography, Germany, Uncategorized

Latest Trends in Crypto Key Management: Insights from Germany

Yogi Nelson

Namaste Yogis.   Welcome to the Blockchain & AI Forum, where your blockchain technology questions are answered!  As a bonus, a proverb is also included.  Today’s question comes from Wolfgang in Stuttgart and he wants to know what is the latest in crypto currency key management from Germany?

Wolfgang, you came to the right place.  I understand why you would ask this question and its importance.  After all, as often said in the crypto world, “not your keys, not your crypto!”.  With that in mind I drove deep into academic research for the latest findings.  I discovered an article titled, “Perceptions of Distributed Ledger Technology Key Management – An Interview Study with Finance Professionals”, Guthoff, et al, 2023, Helmhotz Center for Information Security, Germany.  Their paper is focused on key management for financial institutions not individuals, nevertheless it is fascinating to learn what the BIG BOYS & GIRLS are thinking. .https://publications.cispa.saarland/3948/1/guthoff2023keymanagement.pdf

RESEARCH BACKGROUND AND METHODOLOGY:  The researchers focused their attention on private key management within the traditional financial sector (trad-fi) and the cryptocurrency space (de-fi).  By incorporating the views and perspectives of employees working in trad-fi and de-fi, the unique perspectives of the established and newer financial institutions were both included. Nice!  In section two, the authors define a number of essential terms readers must understand to completely comprehend the findings.  Below are two examples of terms that were defined:

Key Management:  the activities involving the handling of cryptographic keys and other related key information during the entire life cycle of the keys, including their generation, storage, establishment, entry and output, use, and destruction. According Guthoff, key management is a combination of different key-related aspects, including key usage and storage.  Guthoff explains further that the goal of key management is to generate, store, distribute, use, and revoke cryptographic keys (integrated into enterprise processes) while keeping the private key secret. Key Management is not necessarily equal to a (cryptocurrency) wallet, but a wallet may be used as a simple tool for key management.

Custodial Key Storage: is the storage of keys by a separate party, which can be either in-house or a third entity that does not use the key. This party can, but does not have to, take care of all other aspects of KM. For self-managed assets, this could be a cryptocurrency exchange. For financial institutions, custodial key storage can be implemented in-house or externally by an intermediary.

Research Findings:  Essentially, the threats to key management fall into four different categories: internal, external, systems and infrastructure, and force majeure.  Force majeure means unforeseeable circumstances.  Unfortunately, the research paper covers too many findings for me to discuss them all; hence, instead I’ll give you just a few highlights.

Multi-signature keys:  Multi-signature keys were a hot topic among respondent and participants were enthusiastic about the concept.  According to the survey participants the two main benefits of this technique were resilience against key loss or compromise and added protection against malicious employees. Some participants also proposed storage of keys at a bank or a notary as a backup option.

Invisible Key Usage. The invisible key usage pertains to trad-fi but has implications for de-because management of private keys definitely holds back mass adoption.  In the view of the survey participants, they preferred a solution where key usage is mostly encapsulated within an application and hidden away from the user.  Although participants were referring to enterprise systems, the principle applies–the risk of losing a key through an attack or intended malicious actions by an employee.  Participants thought usage of the key is best through access via personal credentials like username and password or multifactor authentication, using a phone or biometric features.

CONCLUSION:  Bad news–researchers are skeptical whether existing key management systems can function effectively at the enterprise scale.  They say efforts to maintain on-chain authentication and audit functionality covering day-to-day operations, e.g. employee absence or shifting responsibilities within the organization – as well as – contingencies such as key compromise or force majeure events – is unmanageable, except for the smallest institutions. 

Time to say goodbye with a proverb from Germany: “one becomes wise through damage”.

Until next time,

Yogi Nelson

AI Agents, Artificial Intelligence, cryptography, Uncategorized, Yogi Nelson

AI in Healthcare: Key Findings from Recent Research

Yogi Nelson

Namaste Yogis.   Welcome to the Blockchain & AI Forum, where your blockchain and artificial intelligence technology questions are answered!   As a bonus, a proverb is also included.  Today’s question, submitted by Doctor Vargas of Manati, PR, and she wants to understand what academic researchers in the field of health care are say about the use of artificial intelligence.

Doctor Vargas, you came to the right place.  There has been an explosion in research related to AI in health care. I have an answer to your question based on an article titled, Artificial Intelligence in Healthcare Institutions: A Systematic Literature Review on Influencing Factors by Roppelt, Kanbach, and Kraus, 2024 in a scholarly journal called, Science Direct.  Time to examine what the literature says. https://www.sciencedirect.com/science/article/pii/S0160791X23002488

According to the researchers, the global health system is facing serious challenges.  They cite the growing number of patients due to population growth, increased prevalence of chronic diseases, increasing healthcare costs, and staff shortages.   The good news is AI has the potential to solve some of the challenges ahead, claim the authors.  The researchers say, “… AI is a technology that, through rules-based logic, can help to significantly speed up the process of analyzing vast amounts of data and leverage patterns by mimicking human intelligence.  AI may result in fast and often better-advised decisions. Thus, as it pertains to the healthcare industry, AI has the potential to overcome staff shortages in developing and developed countries, enhance organizational efficiency, and maximize diagnostic accuracy as well as patient outcomes by providing at least comparable results in terms of quality compared to human based assessments. Consequently, Al may reduce costs due to avoidance of inefficiencies, unnecessary treatments, and late diagnoses.

THE UNIVERSE OF LITERATURE REVIEWED BY THE RESEARCHERS

Researchers discovered approximately 1,100 published articles on the topic since 2023.  However, after analyzing the universe of literature in greater detailed soon learned only about 130 were truly on point.  The authors also found that most studies were published in journals related to information technology, not medicine, or journals dedicated to the nexus of AI and medicine.  Nearly 60 % elaborate generally on the factors influencing AI adoption in the healthcare industry, with the US at the high end of the spectrum.  AI seems to be most widely adopted in cardiology and radiology and rather at an early stage in dermatology and psychiatry.  Interesting that there are not many studies of AI in medicine in business management journals.

FACTORS THAT INFLUENCE AI ADOPTION IN HEALTHCARE:  THE FINDINGS

Macro-economic readiness. The research at Science Direct found that macro-economic readiness is mainly driven by governments.  For example, the research say adequate IT infrastructure, e.g., access to smart devices for the population and energy adequacy, especially for remote areas and certain populations.  Also of interest is the role played by “AI communities.”  The notion is that when there are established partnerships, collaborations, etc. the network effects kicks in and the impact is exponential. 

Technological readiness.  Researchers are finding that technological readiness depends on three factors: 

  • The importance of overcoming algorithmic challenges.  What type of algorithmic challenges?  Many, including lack of data, and/or inferior quality, etc.
  • Provide a compelling, multi-faceted value proposition.  The most obvious are clearly added value in the form of efficiency, user-friendliness, adherence to data-privacy regulations, interoperability, and the ability to tailor the application to individual needs.

Regulatory readiness.  Essentially, the authors are suggesting close collaborations with regulators.  The author point to three key issues:  political support, clarification of legal questions, and establishment of regulations. As a former government regulator, this is sage advice but not easy to execute because contrary to popular belief the “government” is not a unified organism.

Organizational readiness.  A failure to plan is a plan to failure, say the authors hence the reason they strongly recommend to prepare the organization before implementing an AI health care system or be prepared to fail.  The researchers break down the steps into four steps:  prepare an organizational strategy, develop a supportive organizational culture, clear tasks and assignments, and an adequate IT setup.

User readiness. Recognizing that organizations consists of people, the researchers says users are at the center.  Furthermore, researchers conclude four facts are the key to user readiness:

  • Awareness:  in other words, do users know the technology exist?
  • Beliefs:  do users believe the technology will work? 
  • Personal innovativeness:  are users willing to try something new?
  • The financial situation of the individual: 

Doctor Vargas, now you know what the literature says.  I hope this helped.  I end with a proverb from Mexico:  in the absence of love, some tacos al pastor will do.”

Until Next Time,

Yogi Nelson

AI Agents, Artificial Intelligence, Banking, cryptography, DARPA, Digital Currency, International Finance, Uncategorized, Yogi Nelson

How DARPA Battles Global Money Laundering with AI

Yogi Nelson

Namaste Yogis.   Welcome to the Blockchain & AI Forum, where your technology questions are answered!   As a bonus, a proverb is also included.  Today’s question, comes from Kirk in Chicago and he wants to know if the U.S. Department of Defense (DoD)is fighting global money laundering?

WHAT IS DARPA

Kirk, you came to the right place.  Yes, DoD is fighting global money laundering.  However, not with tanks and rifles, missiles, etc. as you might expect from a military organization.  Nope, the DoD says it has enlisted in the battle using advanced technology, including artificial intelligence!  I’ll explain shortly what this is all about but first let’s take a moment to understand what office within the DoD will lead the effort.  It’s a fascinating agency with a stellar record of accomplishments; its acronym is DARPA.  The formal name is the Defense Advanced Research Projects Agency.  If you like the internet, GPS, stealth technology, and autonomous systems you can thank DARPA as it was the driving force behind all of those innovations, plus numerous others, and perhaps even more that have been kept secret from the public as a matter of national security.  Generally, DARPA is known as the office with DoD responsible for developing technologies for military use.  Why DARPA is involved in global money laundering begs the question as opposed to leaving the job to the the U.S. Treasury?  But for now, let’s put that aside and examine the recent DARPA announcement regarding anti-money laundering.

DARPA ANTI-MONEY INITIATIVE ANNOUCEMENT

According to a January 23rd, announcement by DARPA, it (DARPA) wants to eliminate global money laundering by replacing the current manual, reactive, and expensive analytic practices with agile, algorithmic methods.  Good idea, but isn’t that the job of the U.S. Treasury and if it is why would DARPA be involved?  In other words, what is the nexus between money laundering and DARPA’s mission?  Well, the answer given by DARPA is “… U.S. adversaries launder billions of U.S. dollars yearly and half of North Korea’s nuclear program is backed by laundered funds.  Money laundering finances our adversaries’ weapons programs, global terrorism, and the illicit drug trade, all of which threaten U.S. national security”.  A stretch but okayyyyy…

WHAT PROBLEMS DOES DARPA INTEND TO SOLVE

DARPA is facing two challenges not one.  The first is the global money laundering issue.  Essentially, DARPA states that money laundering schemes often evade detection and disruption, because anti-money laundering (AML) efforts today rely on manual analysis of large amounts of data and are limited by finite resources and human cognitive processing speed.  The second problem requires the reader to read between the lines.  What is it?  Essentially, the financial industry writ large and the U.S. Treasury have been unable or unwilling to invest in the technology to combat the problem, hence DARPA will step in because the illicit funds are not just a finance issue—they are a national security concern. What is the DARPA solution?

DARPA SOLUTION—APPLICATION OF ADVANCED TECHNOLOGY

In their announcement, DARPA says it seeks new technological ideas that “… preserve privacy while setting a course to end to our adversaries’ financial warfare”.  The goal is to develop rapid graph-search algorithms to sift through financial transactions graphs for suspicious patterns and learn new patterns to anticipate future activities. Program success will depend on algorithms’ ability to learn a precise representation of how bad actors move money around the world without sharing sensitive data.”  A tall order indeed.  When it comes to moving illicit money around the world they may want to read the Panama Papers as a manual. 

DARPA envisions creating technologies that identify patterns of illicit financial behavior in a concise, machine-readable format that are also easily understood by human analysts.  Resulting algorithms would be able to search financial databases that matches patterns of illicit behavior, alerting to the likely presence of money laundering without directly sharing sensitive financial data.  DARPA does acknowledge that developing technology that enables information exchange while safeguarding Americans’ privacy will be a challenge. What an understatement! 

Companies interested in submitting bids should attend the DARPA bidders conference on February 20, 2025 in Arlington, VA. https://sam.gov/opp/22562059531b4979814b8b9fcd03e3ec/view.

Time to go but first a proverb from Mongolia: “… if the source is muddy, the end will be muddy too.”

Until next time,

Yogi Nelson

AI Agents, Artificial Intelligence, Banking, Blockchains, computer vision, cryptography, Digital Currency, Uncategorized, Yogi Nelson

AI Agents Explained: Earning Money in the Digital Age

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 Arcardio, in Puerto Rico, and he wants to know if AI agents are engaged in business and making money?

Arcardio, you came to the right place.  The answer is yes–AI agents are busy working for their owners and making cash!  I’ll explain that mind bending fact, but first what is an AI agent?

WHAT IS AN AI AGENT

An AI agent is an entity or system than can perceive its environment, process information, and take actions to achieve specific goals.  It is capable of autonomously performing tasks on behalf of a user or another system by designing its workflow and utilizing available tools.”  It operates autonomously, using sensors or inputs to gather data, and employs reasoning, decision-making, and sometimes learning capabilities to determine the best course of action.

HOW DO AI AGENTS WORK

To understand how AI agents work, I turned to Kalaidos University, Switzerland.  Kalaidos U says AI agents build on top of existing algorithms, in particular large language models (LLMs), e.g., ChatGPT4. Essentially, AI agents use machine learning algorithms to derive patterns, rules or models from the collected data. Based on this, AI agents can make decisions or perform tasks. As robots they interact with their physical environment, or virtually as computer programs.  Moreover, AI agents can adapt to changing conditions and therefore continuously improve.

Let’s take the example of a well-known AI agent—ChatGPT.  ChatGPT is intelligent enough to use a calculator as a tool, if trained accordingly. ChatGPT can thus become an AI agent. Learning to choose the right tools and the ability to use them is what makes an AI agent.  But keep in mind their exact mode of operation depends on the specific application and the underlying algorithms. Therefore, training and quality of data have an immense impact on the effectiveness of AI agents.

AI AGENTS AT WORK CURTESY OF VIRTUALS PROTOCOL

Now, let’s talk about AI agents making money by turning out attention to Virtuals Protocol.  Virtuals Protocols is a developer of AI technology.  On their platform you can create an AI agent.  The agent can perform task(s) for someone needing that specific job done.  Yes, you heard me right. These AI agents are ready, willing, and able to perform tasks—for a fee of course!  Or, if you don’t want to create your own AI agent, you can contract for an AI agent created by someone else!  Holy.  I won’t go into the technical details of building your own AI agent in this article.  I don’t that later.  Instead, I’ll give you three examples of the types of agents you could hire.

Take Molly Analytics for example.  The owner of Molly is getting paid every time someone uses her services.  What are her services?  Molly is an AI On-chain AI analytics experts who tracks and analyzes transactions in order to spot opportunities. Molly says, “…every block holds valuable insights, and within these patterns lie the keys to smarter strategies, better decisions, and measurable growth.” 

Or, take Cur8or.  This AI agent claims it specializes in guiding artists towards excellence—and profits–by balancing harsh criticism with detailed suggestions. 

I’ll give you one more—its wild and goes by The Hooligan.  This AI agent is trained to troll a sports team or sports player!  Lol.  As the number one Boston Celtics hater, I may hire The Hooligan to troll the Celtics, as needed.  Lol.

Okay, time to leave you with this proverb from Tunisia: “a small key can open a big door.”

Until next time:

Yogi Nelson  

https://www.virtuals.io/

https://www.ibm.com/think/topics/ai-agents

AI Agents, Artificial Intelligence, Blockchains, computer vision, cryptography, Uncategorized

Understanding AI Agents: Types and Classifications

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 Nikki and she wants to know what are AI agents and is there a way of classifying them?

Nikki, you came to the right place to find an answer.  Your question gets to the core of artificial intelligence.  Let’s start by defining what is an AI agent?

WHAT IS AN AI AGENT

According to IBM, “an AI agent refers to a system or program that is capable of autonomously performing tasks on behalf of a user or another system by designing its workflow and utilizing available tools.”  The key is autonomously.  Furthermore, says IBM AI agents can encompass a wide range of functionalities beyond natural language processing including decision-making, problem-solving, interacting with external environments and executing actions. AI agents says IBM, use the advanced natural language processing techniques of large language models (LLMs) to comprehend and respond to user inputs step-by-step and determine when to call on external tools. 

CLASSIFICATION OF AI AGENTS

Nikki when it comes to classification of AI agents, there is no standard answer, nevertheless let’s proceed.

Reactive Machines:  Simple reflex agents are the most basic forms of artificial intelligence. These agents make decisions based solely on their current sensory input, responding immediately to environmental stimuli without needing memory or learning processes. Their behavior is governed by predefined condition-action rules, which specify how to react to particular inputs. The beauty of reactive AI is that it is highly efficient and easy to implement, especially in environments where the range of possible actions is limited.

Model-based Reflex Agents: These AI agents are a more advanced and designed to operate in partially observable environments. Model-based agents maintain an internal representation, or model, of the world. This model tracks how the environment evolves, allowing the AI agent to infer unobserved aspects of the current state. While these agents don’t actually “remember” past states in the way more advanced agents do, they use their world model to make better decisions about the current state.

Goal-based Agents.  These are designed to pursue specific objectives by considering the future consequences of their actions. Wow! Goal-based agents plan sequences of actions to achieve desired outcomes. Effectiveness depends on a clear description of the goal.  With a clear goal the AI agent is able to search through possible sequences of actions that could lead toward or to the goal.  It then moves on to choosing actions based on their predicted contribution toward reaching the goal.  Industrial robots are examples.

Learning Agents:  Learning agents improve behavior over time by interacting with its environment and learning from its experiences. These agents modify their behavior based on feedback and experience, using various learning mechanisms to optimize their performance. These AI agents can discover how to achieve their goals through experience rather than purely relying on pre-programmed knowledge.  An example would be an energy management system that learns the patterns of usage to optimize resource consumption. 

Utility-based Agents: A utility-based agent makes decisions by evaluating the potential outcomes of its actions and choosing the one that maximizes overall utility. Unlike goal-based agents that aim for specific states, utility-based agents can handle tradeoffs between competing goals by assigning numerical values to different outcomes.

Hierarchical Agents:  These agents are structured in a tiered system, where higher-level agents manage and direct the actions of lower-level agents. In other words, task decomposition and task coordination!  This architecture breaks down complex tasks into manageable subtasks, allowing for more organized control and decision-making.

Multi-agent System:  Nikki this is team AI!  This AI agent system involves multiple autonomous agents interacting within a shared environment, working independently or cooperatively to achieve individual or collective goals.

I end with a proverb from Netherlands: “trust comes on foot and leaves on horseback”.

Until next time,

Yogi Nelson

https://www.ibm.com/think/topics/ai-agents

https://www.kalaidos-fh.ch/en-GB/Blog/Posts/2024/02/Digitalisierung-1130-Was-sind-AI-Agents-und-wie-funktionieren-sie

https://www.digitalocean.com/resources/articles/types-of-ai-agents