Unleashing AI: Paving the Way for the Future

Unleashing AI: Paving the Way for the Future

"Every technology shift is an opportunity to advance scientific discovery, accelerate human progress, and improve lives. I believe the transition we are seeing right now with Artificial Intelligence (AI) will be the most profound in our lifetimes, far bigger than the shift to mobile or to the web before it. AI has the potential to create opportunities — from every day to the extraordinary — for people everywhere. It will bring new waves of innovation and economic progress and drive knowledge, learning, creativity and productivity on a scale we haven’t seen before...." by Sundar Pichai, CEO, Google and Alphabet

HI and AI

While the goal of AI is to build and create intelligent systems that are capable of doing jobs that are analogous to those performed by humans, we can't help but question if AI is adequate on its own.

Human intelligence (HI) is the ability of humans to think, learn, and reason. It’s the ability to make decisions based on information and understanding. This means that human beings are able to process complex information in a way that computers cannot.

Some experts argue that AI will eventually surpass human intelligence as computer systems become more sophisticated and able to replicate or even exceed human cognitive abilities.

AI vs ML vs DL

Large Language Model (LLM)

A large language model is a type of artificial intelligence algorithm that uses deep learning techniques and massively large data sets to understand, summarize, generate and predict new content. The term generative AI also is closely connected with LLMs, which are, in fact, a type of generative AI that has been specifically architected to help generate text-based content. Over millennia, humans developed spoken languages to communicate. Language is at the core of all forms of human and technological communication; it provides the words, semantics and grammar needed to convey ideas and concepts. In the AI world, a language model serves a similar purpose, providing a basis to communicate and generate new concepts.

Generative AI (GenAI)

GenAI refers to AI systems that create new content, such as text, images, music, or code, by learning from large datasets. It uses models like neural networks to identify patterns and produce human-like outputs. GenAI is closely linked with Large Language Models (LLMs), which are designed to understand and generate human-like text. LLMs, such as GPT-4, are essential for applications like chatbots and automated content creation. GenAI improves productivity by automating repetitive tasks, enabling faster content creation, and enhancing data analysis, allowing employees to focus on strategic decision-making and fostering innovation.

The Rise in LLMs - OpenAI's like ChatGPT, Microsoft's Copilot and Google's Gemini

LLMs are a type of AI capable of generating human-quality text. Today's leading chatbots for the everyday user — OpenAI's ChatGPT, Microsoft's Copilot and Google's Gemini — are all best approached as tutors or assistants rather than information desks.

Why it matters: Getting started using any of these tools is easy but knowing when and how to get the best results from them can still be tricky.

The big picture: All three of these chatbots are designed to create, edit and summarize content.

  • ChatGPT works well for most tasks. Google's Gemini has access to more timely and updated information and Copilot is best for work tasks.
  • Chatbots can help you write your resume, create a basic recipe with the ingredients you have in your fridge, summarize long articles, turn text into presentations, or create itineraries for vacations.
  • They're also great for sparking creativity or combatting writers’ block. Some people even use them to help communicate with others.

Between the lines: Most free versions of these chatbots are good enough. But if you want to take advantage of more features, you can subscribe to any of them for around $20 per month.

  • Copilot is ChatGPT, but with access to all the data in your calendar, emails, chats, documents, meetings, and everything else that's known as your Microsoft Graph (if you use Microsoft 365).
  • Gemini Advanced will let you summarize content or create new content from your Gmail, Google docs, and your Google Drive. You can try it free for two months.
  • While ChatGPT has its own GPT store, Copilot lets you integrate plug-ins from third-parties like Shopify, OpenTable, or Kayak, while Gemini allows you to use extensions like Google Flights and Google Hotels.

How it works: You can access ChatGPT, Copilot, and Gemini on the web.

  • ChatGPT and Copilot both have free iPhone and Android apps, but Google's Gemini is only available as an app on Android. You can access Gemini on iOS through Google app by tapping the Gemini icon.
  • The request you make or question you ask Copilot, Gemini, or ChatGPT is called a prompt.
  • Prompts should be specific and can include requests for responses in a particular tone, a certain audience, or even a specific number of words.
  • For example, "Explain the theory of relativity to me in 100 words. Act as a friendly scientist and answer as if I'm ten years old."

Fun fact: Studies have shown that you'll get better results if you're polite with the tools.

Reality check: All current generative AI chatbots are prone to producing inaccuracies and historical untruths, known as "hallucinating."

  • The web has been full of misinformation since its inception, but search engine algorithms have been getting progressively better about showing us more reliable sources.
  • Chatbots are setting us back a few steps since by their very nature they're designed to make stuff up.
  • The confidence with which these chatbots will deliver misinformation among other more reliable information can be disconcerting.
  • Any prompts you enter could potentially be used to train these chatbots, which means you should be extra careful with sensitive personal details and confidential work information. If you pay to subscribe, some services will do more to respect your privacy. Google says your personal content from Workspace won't be used to train Gemini's model or viewed by human reviewers even in the free version.

What's next: Soon AI will not only be able to answer questions and create or summarize content, but it will also be able to act on your behalf — which is both exciting and terrifying.

Improve your work with AI tools

ChatGPT

ChatGPT from OpenAI is easy to use. To use the tools from OpenAI, you'll need an account. If you don't have one, simply register with your Google or Microsoft account - it's that easy! Like other apps you can use this OpenAI tool in your PC and mobile.

OpenAI's ChatGPT tool is designed to give detailed responses to any inquiry you type - from questions to statements. While the best results come from inputting a statement, questions are also accepted. Specific instructions lead to more accurate and specific answers from ChatGPT. To ensure accuracy, you can input queries continuously until you close your browser or reset the thread to clear your previous requests. You can also choose to view the page in either dark or light mode.

To create account, you can register here.

User acceptance and popularity

Here we can see the active growth status of some apps. Where we can see ChatGPT is the fastest growing app in terms of user acceptance. Reached 1 million users in 5 days only.

ChatGPT, powered by OpenAI’s GPT-4, has significantly improved in accuracy and breadth of knowledge. However, it’s still not foolproof; when there isn’t enough data available, it may occasionally fill in gaps with general information. OpenAI has addressed many limitations present in earlier versions GPT 3 and 3.5 but notes that the tool's knowledge of events after its last training cut-off in 2023 might still be incomplete.

As of the last update in June 2024, OpenAI does not enable real-time internet lookups for ChatGPT in its standard configuration, primarily to ensure consistency and reliability in responses. This means that while the capability might exist, it is not utilized in the regular, publicly available version of ChatGPT. However, it’s always good to check the latest from OpenAI directly, as features and capabilities can evolve.

How to utilize AI in your workplace

Summaries

You can use AI to summarize a document for you. You can give it a document and ask it to summarize it for you. This is a very useful feature for people who need to summarize documents quickly. This also comes in handy when you are researching some topic. Just paste long articles into AI and it will summarize them for you. Additionally, you can use AI to summarize the entire website and ask questions about the text. This makes it super-fast to extract information like phone numbers, addresses, etc. from websites.

Emails

You can use GPT-4 to help you with your emails by providing it with some information about what you want to say in your email, and it will generate text that you can use in your email. For example, you can give GPT-4 the topic of your email and some bullet points that you want to include, and it will generate the body of the email for you. You can then review and edit the text as needed.

Brainstorming

You can use AI to help your brainstorm ideas. You can give it a topic and ask it to generate ideas for you. It will then generate a list of ideas that you can use in your work. It's possible to ask to elaborate on certain key points and AI will talk a little more about them. This is amazing to use when you are stuck on a project or assignment and need some inspiration.

Apart from the above email, summaries and brainstorming for idea generation, we can also use AI to Proofread and Rephrase our text to reduce error in writing and have more professional writeups. You can even request changes to the text or use regenerate option, and AI will generate new text that is more similar to what you want to say.

As we have seen ChatGPT is popular according to user acceptance as above. In line with Google Gemini and Microsoft Copilot comes with their AI for the users to add some advancements in their work. The user interface of Google Gemini and Microsoft Copilot are almost the same as ChatGPT with some other features added.

Microsoft Copilot is primarily meant to support Microsoft 365 users with automation features for Word, Excel, PowerPoint, Outlook and Teams. It also works with the Bing search engine and Windows operating system (OS). Google Gemini, a multimodal AI by DeepMind, processes text, audio, images, and more. It includes three versions: Ultra, Pro, and Nano, each tailored for different task complexities. One can also choose any of these AI tools as an AI assistance at work. One can get so many online free tutorials related to the usage of those AI tools.

How AI Can Help Internal Auditors in Drafting Audit Reports

AI can be a valuable tool for internal auditors when preparing audit reports by streamlining the drafting process, enhancing clarity, and providing insights. First, AI can assist in gathering and synthesizing relevant data by analyzing large volumes of documents and extracting key information. This saves significant time for auditors on the research phase and allows auditors to focus on analysis. Some of the assistants of AI are-

  • Drafting Assistance: AI can help auditors quickly draft sections of the report on given templates, offer well-structured language for findings, analyzing suggested risks and recommendations.
  • Time Efficiency: By automating repetitive writing tasks, AI allows auditors to focus more on analysis rather than writing.
  • Clarity Enhancement: t can review and improve the tone and clarity of the report, translating technical jargon into clear language for stakeholders.
  • Fact-Checking: AI can assist in checking industry compliance standards and best practices to ensure accuracy.
  • Simulating Responses: It can simulate responses to hypothetical queries from audit committees, helping auditors anticipate potential questions and concerns.
  • Tailored Insights: The tool can offer insights and suggestions, but auditors should verify content and ensure compliance with organizational guidelines.

However, auditors should ensure that any generated content aligns with their specific organizational guidelines and verify all data for accuracy.

Concern matters of using AI

AI tools have become invaluable in the workplace, streamlining tasks and enhancing productivity. They assist in data analysis, customer service, and even creative projects, offering efficiency and innovation. However, AI also has its drawbacks, such as the potential to disseminate inaccurate or misleading information. This underscores the importance of human oversight to ensure the reliability and integrity of AI-generated content.

Caution for auditors in using AI

Caution for auditors in using AI

Some caution steps auditors should follow when using AI in performing audit tasks:

  • Data Confidentiality: Never input sensitive or confidential client data into AI. Example: avoid sharing internal financial statements, PII (Personally Identifiable Information) e.g., name, address, NID / smart ID number or other identifying PIN or code, telephone number, email address, etc. and at organizational level e.g., company name, staffs name, designation, ID or PIN code, organizational policy & procedures etc. to prevent data leaks.
  • Verification of Content: Always verify the accuracy and relevance of content generated by AI before including it in reports. Example: cross-check any regulatory or industry standard suggestions provided by the tool with official guidelines to ensure they are up to date.
  •  Customization for Context: Ensure the generated content aligns with the specific audit context or operations. Example: if AI provides a generic audit template, auditors should tailor it to meet the specific requirements and nuances of their audit process.
  • Critical Thinking: Avoid over-reliance on the tool. Auditors should still apply their professional judgment and skepticism in all audit matters. Example: if AI drafts a recommendation, an auditor should assess whether it is practical and aligned with the company’s risk profile and current practices.
  • Legal and Compliance Risks: Ensure that the output from AI complies with relevant legal, regulatory, and ethical standards. Example: before incorporating any content, check whether it aligns with jurisdictional laws or industry-specific regulations, like the Data Protection Act (DPA), Labor Act, data privacy policy of the organization and code of conduct etc.
  • Tool Limitations: Recognize that AI lacks industry-specific expertise and may not always reflect the latest developments in audit standards. Example: While AI can suggest audit frameworks, it may not account for recent changes in auditing standards such as updates from the IIA (Institute of Internal Auditors), IASB (International Accounting Standards Board) or IFAC (International Federation of Accountants).

Recently, AI-generated deepfake videos have caused significant concern, illustrating the potential for technology to create highly convincing but false representations of real people. For instance, deepfakes of prominent political figures have been circulated online, leading to misinformation and public confusion. Additionally, the use of AI in facial recognition technology has raised privacy issues, with reports of wrongful arrests due to inaccurate identification. These incidents highlight the critical need for robust ethical standards and regulatory measures in the deployment of AI technologies.

Case Study:

The Impact of Fake Pentagon Explosion Images Created by AI

Key Takeaway

This article emphasizes the transformative potential of AI in enhancing productivity and innovation in professional settings. The evolution of AI, particularly in chatbots and language models, brings both advancements and challenges. While AI tools like ChatGPT offer valuable capabilities for summarization and response generation, there is a need for vigilance in detecting misinformation and understanding the limitations of AI systems. Users should leverage AI responsibly, be cautious with sensitive information, and stay informed about the latest developments in AI technology and its implications.

This CPD article can be simply concluded with below two sentences -

The world is moving towards AI, where you are standing ?Adhere to Jack Welch's words, “Change before you have to”



If you have any queries, please feel free to reach out to me on my LinkedIn linked below.


Prepared by:

Kazi Tripti Hasan

Internal Audit Department, BRAC

Linkedin ID: linkedin.com/in/tripti-hasan-09b63955


Zahirul Jewel, ACCA

| MSc, UK - AACSB Accredited | Diversified Industry Experience | Finance Operations | Financial & Management Reporting |

2w

Insightful article.

Snehasish Barua

Partner at Snehasish Mahmud & Co

3w

Keep it up

Sk Md Tarikul Islam

Partner at Hoda Vasi Chowdhury & Co, FCA (ICAB), ACA (ICAEW), MBA (Bradford, UK), Chartered Business Valuator (CBV-Canada)

3w

Good one

Very good job Tripti apa. Thank you for sharing.

Md. Zahed Sharif

AGM at BRAC | Internal Controls | Risk Management | Fraud Management | Compliance | Data Analyse | CA CC | DU | NDC

3w

Very informative and very helpful. Thank you for sharing this with us.

To view or add a comment, sign in

Insights from the community

Others also viewed

Explore topics