From Sci-Fi to Reality: The AI Evolution

AI and the Human Experience: "Enhancing Our Lives"

1. Introduction:

The term, concept, phenomenon, now commonly referred to as "AI" (Artificial Intelligence), has received significant interest from large, medium and small organizations, investors and even entry level to executive level corporate and non-corporate employees and entrepreneurs. The rate at which AI models and tools have developed across the global markets has led us to the point, where, defining "AI" is somewhat impossible. However, let's give it a shot.

"Artificial Intelligence (herein after referred to as "AI") refers to the theory and development of computer systems that can perform tasks traditionally requiring human intelligence, such as recognizing speech, making decisions, and identifying patterns. AI includes various technologies, such as machine learning, deep learning, and natural language processing (NLP)."

World renowned entrepreneur, Elon Musk, owner of Tesla and X (formerly Twitter) predicted the imminent rise in AI superintelligence during a live interview and further went on to say; "My guess is we'll have AI smarter than any one human probably around the end of next year".

Despite significant criticism to his averments that AI would exceed Human Intelligence, this has in no way deterred or constrained his views and ambitions around the development of AI. Musk has recently launched another AI focused company/initiative named or known as "XAI", formally known as "Explainable AI", the primary purpose of this company/initiative is to ensure that AI is both, understandable and transparent to humans, a key focus of the XAI initiative directly correlated with the increasing uncertainty around the ethical use of AI and regulations aimed at benefitting humanity.

A few of the most sophisticated AI platforms available on the market today are:

• ChatGPT
• Claude AI
• Google Gemini
• Midjourney
• Microsoft Copilot

2. Generative AI:

The development of Generative AI tools is certainly an area of AI that I am particularly fascinated with. I recently completed a learning path focused on Generative AI wherein significant insights were shared, including, but not limited to, the "Generative Adversarial Networks (GANS), the "Generator vs Discriminator" effect, among many others. The Learning Path additionally highlighted and elaborated on the differences between Reasoning Engines and Search Engines, an important distinction to make prior to engaging/using AI tools.

3. Mastering the Art of Prompt Engineering:

Drafting prompt's, a critical and underrated skill. Over the course of the last year, I have endeavored to master the art of drafting prompts through the completion of multiple AI courses, online learning and independent research as well as practical application of the skills/techniques that I have acquired. A few of the key considerations as your attempt to create/draft the "perfect" prompt are:

• Be specific, provide context, elaborate on real life scenarios, provide examples using clear and concise language.
• The use of role-play scenarios, analogies, debate style questions (for and against arguments).
• "Meta prompting" - utilization of the AI tool to suggest possible prompts or improve your initial prompt/s in order to achieve the desired outcome.
• Provide guidelines (formatting suggestions, define scope, identify and state the goal or purpose you are aiming for).
• Include Statistics and Anecdote.

The list above is in no way exhaustive but merely a few examples for AI users to improve their prompt drafting skills, through personal experience, I can confirm that the phrase "practice makes perfect" could not be more suitable in this instance.

4. Synopsis:

The AI Tool I have primarily and most frequently used is Microsoft CoPilot. I have been fortunate enough to gain access to and use of Copilot for MS 365 for Business Users wherein there is interaction with the organizations Microsoft 365 data (paid subscription).

The interaction between the organizations data and the AI Tool (CoPilot) have proven to be of significant value, allowing for faster responses, access to newer language models as well as the ability to use CoPilot within Microsoft Applications such as MS Outlook, MS Teams, MS Excel, MS PowerPoint and MS Word (to name a few).

A few examples of these added benefits include:

MS Teams

• Summary of Communications (purpose of the meeting, primary discussion points)
• Ask specific questions to provide information within chat to answer questions (generate a summary/draft of the meetings topics).

MS Outlook:

• Summarize, draft or rewrite email using Copilot.
• Summarize chain of email communications.

MS PowerPoint:

• Create presentation
• Transform a file – insert/export transcript from a MS Word Document
• Ability to export script from an MS word Document or by coping the VBA code from the CoPilot Script to create a presentation with the added benefit of automatically creating the presentation on a company/corporate defined/standard template.

MS Excel:

• Improved efficiency in use of common/frequently used functions
• Sort and Filter data
• Combine sorting and filtering
• Analyse datasets
• Automate creation of pivots tables and charts
• Suggestions to use formulas and correction of inconsistencies

MS Word:

• Select text – auto rewrite
• Choose tone to be used
• Reference a stored file

5. Establishing an Ethical AI Framework:

Responsible Data Practices:

• Analyze, review and update source of training data.
• Reduce risk of implicit or explicit bias.
• What measures are we using to prevent bias decision making in future?

Privacy:

• Conduct an AI Focused Privacy Audit.
• Privacy Policy in line with Global Legislative Frameworks (GDPR, DORA and more specifically the EU AI Act, which will regulate the use of Artificial Intelligence within the EU, thus, making it the world's first comprehensive AI law recently mentioned in an article titled "Conquering Cross-Border Complexity with Effective Risk Management") to protect confidential data and personal information.
• Adequate and frequent training.

Reduce Bias:

• Bias Audit.
• Does data represent population you are serving?
• Re-training of model if subject to analysis of outcomes?
• Are we asking the right questions?
• "Cross referencing", a technique used to eliminate the potential risk of bias through the use of a few multiple accredited and reputable AI tools in order to conduct a comparative analysis of the outcomes and subsequently identify where or what element of the dataset is the root cause of the bias and how it can be eliminated.

Technology Teams – skills and expertise

• Internal Ethical Culture - build a strong culture around ethics and responsible practices even within and specifically within teams/workstreams who's area of focus is typically occupied by individuals who perceive stringent adherence organizational policies and guidelines as a "blocker" as opposed to a form of protection and long-term enabler of sustainable and efficient workstreams.
• Security and privacy of data collection.
• Audit algorithms ensuring they are fair and avoid bias outcomes (criteria to be defined and assessed periodically to ensure elimination of risk).
• Ethical Training for Technology Teams serves a dual purpose, the provision of training around the use of AI which will be the Status Quo for all employees as well as emphasizing and clearly communicating the role the Technology Focused Teams have to play in the endeavor of creating a framework suitable for the organizations needs and in line with global standards.

Executive Team Directing Ethical AI Framework:

• AI Policy and Governance Framework, a task or requirement that could be outsourced to a specialized global law firm or consulting firm to expedite the process of analyzing the legislation and generating the framework, alternatively, control and longer-term investment given the increased usage and risk associated with the use of AI - hiring or creation of an AI focused team focused on the development of the AI Framework could prove to be of more benefit to the organization, ideally if part of the organizations strategic goals is to increase the use of AI tools across the organizations warrants the need for an AI focused team.
• Adherence to Data Privacy best practices.
• Democratizing decision-making regarding AI Policy.
• Defining and Directing AI Focused Audits.

6. Conclusion:

The more I learn about the evolution of AI, the foundations upon which AI was built as well as the practical application and use of AI Tools, I am significantly more optimistic about AI. The benefits of effectively and responsibly utilizing AI across and within various industries and disciplines significantly outweigh any radical transformation/evolution that we have witnessed over the last 3 decades. There are multiple contradicting views and misconceptions around AI such as "the replacement of humans by AI, reduced employment opportunities for humans, the list goes on". AI should be seen or perceived as an assistant to humans/humanity in a number of ways (many of which we have most likely not even thought of or predicted). The framework, implementation and practical application of any proposed or approved legislative changes and/or increased regulation and control mechanisms around AI usage should not be perceived as a threat to humanity but rather a benefit enabling humanity to not only enhance operational efficiency, generate ideas, but also challenge conventional ways of thinking, improved speed and accuracy of analyzing data thus enabling organizations to perform predictive analysis and forecasting in a robust fashion directly related to the "Evolution of Data" and Platforms with built in integrations, user configuration, API authenticated creating a seamless flow of data into a structured data set/ data warehouse / data hub.

A personal ambition of my own, as a Senior Associate in the Alternatives Investment Funds Industry, an area that I have identified and purely initiated on my own accord is conducting research, industry specific research, further divided by region, location, service offerings, competitor analysis, gap analysis, identifying and monitoring market trends (inflation rates, interest rates, new/amendments to legislative frameworks, enhanced due diligence) to name a few.

7. Leveraging AI: "The Jury is still out on this one"

• Credible External Data Providers such as Preqin, Pitchbook and Bloomberg publish and release quarterly and annual reports, these reports are further divided into subsectors (regional reports, asset class/industry specific reports).
• Internal Data - Identification and structuring of our organizations internal data such as our operational metrics (timeliness, quality, resourcing, profitability, number of clients we service, geographical coverage, the type of clients we service, the scope of our service offering, the number of Errors, Losses, Delays or Complaints as well as the reasoning for the delay or error and the following remediation process).
• Data is divided into two primary categories: Quantitative Data (Statistics, values, numerically measurable data) and Qualitative Data (reason for errors, reasons for decline or increase in certain locations or regions due to a variety of factors such as market uncertainty, economic risks, high inflation rates).
• A holistic view of the dataset is split into two categories: External Data Providers and Internal Data from within our organization
• These two datasets are further divided into two sub-categories: Quantitative and Qualitative Data, with due consideration to instances where certain data points overlap both sub-categories in one way or another.
• Developing a sophisticated inventory of prompts (subject to change and iteration continuously) aimed at extracting the identified and critical quantitative and qualitative data from external data providers in cohesion with an inventory of prompts to extract the identified and required quantitative and qualitative data from within our internal dataset.
• Drafting and developing a prompt or a number of prompts in order to leverage AI to not only extract this data but consolidate it into a report/executive summary which can be used to make decisions that affect our internal operations as well as provide insights in a robust, structured and accurate way to our clients, thus, truly embracing the ideology of "Data-Driven" decision making.

"AI is the most important thing humanity has ever worked on. I think of it as something more profound than electricity or fire." - Sundar Pichai, (CEO, Alphabet Inc. and its subsidiary Google)

8. Bibliography:

• What Is Artificial Intelligence? Definition, Uses, and Types | Coursera
• Elon Musk: AI will be smarter than any human around the end of next year - Ars Technica
• What is XAI? Elon Musk's Vision for AI and His New Project
• https://www.garudax.id/learning/certificates/b9dabda529867ca7503efe84650222ebfd2aee6f73bb6c06454d09e0bbfebfee?trk=share_certificate
• https://www.garudax.id/learning/certificates/b9dabda529867ca7503efe84650222ebfd2aee6f73bb6c06454d09e0bbfebfee?trk=share_certificate
• https://www.garudax.id/learning/certificates/c77e4daace588c721ecf1abedd2124dcfd4cd519e0326a77617e4501c7fdde60?lipi=urn%3Ali%3Apage%3Ad_flagship3_profile_view_base_certifications_details%3BPo1Z47F7SH6eJWMJ0QnTdg%3D%3D
• https://www.garudax.id/learning/certificates/12ff632604e62e15c0f8ad992f61c49d778b71577b18591bc78075cd43c408b5?lipi=urn%3Ali%3Apage%3Ad_flagship3_profile_view_base_certifications_details%3BPo1Z47F7SH6eJWMJ0QnTdg%3D%3D
• https://www.garudax.id/learning/certificates/56729c38178708e1421d4706c0521e8f52d10f0f3af33625e60dfe97182db049?lipi=urn%3Ali%3Apage%3Ad_flagship3_profile_view_base_certifications_details%3BPo1Z47F7SH6eJWMJ0QnTdg%3D%3D
• Conquering Cross-Border Complexity with Effective Risk Management Aynsley Vaughan