top of page
Search

Reflecting on My Journey with Generative AI



I've been using generative AI for about over a year now, both for personal and professional tasks. It's quickly become a key tool in my daily life, producing a level of excitement I've never felt before. It's like having Aladdin's genie at my fingertips, always ready to discuss any topic or assist with tasks. I'm impressed by how well it seems to understand me, often providing the exact answers I seek. Proper prompt engineering plays a crucial role in communicating effectively with systems like GPT-4 and beyond. Frankly speaking, It has given wings to my thoughts, allowing them to fly high in the sky.

Since childhood, I've pondered various deep questions, such as our purpose on life, the mysteries of the universe, and the nature of existence etc. This AI has allowed me to delve into these topics even further. It feels like the GenAI Assistant, and I have built a strong bond over these months. There's so much I've accomplished with its help that I'm eager to share, from personal insights to its impact on my work as an architect.

When I discovered the power of code interpreter feature in Chat GPT Plus, I couldn't stop myself to have a monthly paid subscription of it. I will cover the great feature code interpreter (beta) in a separate post but in short it is like putting AI at work, you just give it a work (text data, audio or video) with proper prompt and it gets that work done for you with high quality in quick time. It can generate data for you as well and lot more to offer.

My POV on Generative AI, Foundation Model, Large Language Model and Prompt Engineering...

I'm convinced that advancements like GenAI/LLM/FM will become foundational in shaping our future, both on Earth and beyond. These technologies have the potential to revolutionize the way we operate and interact.


Across industries, there's an unyielding drive to harness AI's immense capabilities to spearhead business growth and maintain a competitive edge. On an individual level, many are eager to tap into AI's promise to elevate their day-to-day experiences. AI's expanding influence suggests it's on track to become an integral part of our global fabric, offering vast opportunities while posing significant challenges.


It has superpower energy that has high potential of giving exponential benefits. As we know with great power comes great responsibility. So, applies here as with greatest power comes with greatest responsibility. This marks the beginning of a new era where humanity can either ascend to unprecedented heights or descend into destructive lows. It is like it possesses an immense reservoir of energy, similar to atomic power, which can either be harnessed for great advancements or misused to catastrophic consequences. Just as controlled use of atomic energy is crucial, so too is the prudent and ethical use of AI to ensure a positive and beneficial impact on our world. Therefore, exercising caution, responsibility, and wisdom in its application is of utmost importance.
As an example, we've already seen a little over excited use of it, made google a billion-dollar loss recently. We (IBM) have to learn from it and not to repeat the similar. We also heard Samsung stories and I believe there are many undiscovered yet.
Below are my understanding of these new famous terms of modern AI based on my experience in recent times.

Generative AI

Generative Artificial Intelligence is more than just a tool for content creation; it's a powerful analytical engine capable of generating new insights based on vast amounts of data. Generative Artificial Intelligence (GAI) stands at the intersection of machine learning, creativity, and data analysis. It's application in business and life on earth and beyond is unlimited.


Foundation Model

The term "foundation" in "foundation models" reflects the idea that these models serve as a foundational layer or starting point for a wide array of applications and tasks. A foundation model is pre-trained on vast amounts of data, allowing it to capture a broad understanding of language, images, or other modalities, depending on its training data. This broad understanding serves as the "foundation" for further customization or specialization.


Large Language model

A "Large Language Model" is called "large" due to its vast number of parameters and "language" because it's specialized in processing human language. Large Language Models like GPT-3 or GPT-4 are foundation models trained on vast amounts of diverse internet text to capture a broad understanding of language. It's trained on vast amounts of text to recognize patterns, allowing it to answer questions, write essays, help with translations, and perform other language-related tasks. It is like the most knowledgeable machine/model as of now.


Prompt engineering

You're speaking to an advanced machine, similar to interacting with an expert. Just as you'd adjust your conversation style based on someone's expertise, you should do the same with AI. Prompt engineering is a way to guide AI systems like ChatGPT. It helps you get the answers you want without changing the AI itself. The better your questions, the better the AI's answers. Plus, clear questions mean shorter, precise answers, saving you time and system resources, especially if you're using the AI for work tasks.


Balancing Natural Intelligence and Artificial Intelligence

In our modern era, the comparison of Natural Intelligence (NI) and Artificial Intelligence (GenAI) is more relevant than ever. With the advancement of AI models that can generate content, be it text, images, music, or even novel ideas, the balance between human cognition and machine-driven creativity is crucial. This balance can unlock unprecedented potential, especially when we understand the limitations of human intelligence and the areas where GenAI can be most impactful. Soon I am coming up with a separate post on this.


GenAI (Genie Lamp) as my Professional Assistant

As a senior architect persona, prompt engineering has been invaluable in my various tasks since it's launch. It assists me in learning new tools, generating innovative ideas, designing architectures, helping in taking architectural decisions, code assistants, generating documents, comparing products and services, creating presentations, analyzing data, summarizing requirements, writing emails and blogs, and exploring more use cases and so many things. Soon, I'll share the prompts I've used, highlighting which ones worked exceptionally well or just partially. There are many I saved them. Some of the work is like a magic.


GenAI (Genie Lamp) as my Personal Assistant

On a personal note, I've been using it to seek answers about life and the universe, assist my child in preparing for IIT/engineering exams, familiarize myself with new tools and technologies, and delve into hobbies like photography and music & singing. It's also been handy in learning about guitar techniques, health best practices, and weight loss etc.


Cautions & Best Practices

We're dealing or playing with a super powerful systems hence extra cautions are required. It is like a fire that can be used for benefits or can burn you if not used appropriately. The promise of GPT-powered AI is undeniable, but with great power comes great responsibility. It's upon us — developers, users, and regulators — to ensure that we harness this potential in ways that benefit society at large while minimizing potential harms. There are many best practices. Few of them mentioned below that can help

  • Learn prompt engineering best practices. Be aware of the model's capabilities and limitations.

  • Clarify your role and the role and functions of the GenAI tool during discussions on various topics.

  • Apply critical thinking and problem-solving when engaging with AI systems. Stay attentive and avoid being casual. Relying on common sense can be beneficial. Clearly identify your objectives when seeking assistance from AI.

  • Provide a clear and specific prompt to achieve the best results. It's important to choose the right context when using GenAI tool.

  • Ask information with backed by sources.

  • Don't share private and sensitive information of any entity to public AI systems like chatGPT etc. Instead mask it if it really needs it to. If google search is allowed in your organization, you may consider to use like that.

  • Don't blindly use the response output as it is. Read it and understand it, improve it as per your persona and make it sense and connection until your maturity mind satisfies and then use it.

  • You may use multiple tools to ensure you're getting similar results and are correct.

While advanced AI systems bring tremendous potential, they also pose concerns related to misinformation, biases, security, and broader societal impacts. Measures like refined prompt engineering, robust governance, targeted education, and increased transparency can significantly address and mitigate these issues. However, ensuring the ethical and safe deployment of AI is an ongoing effort, and we must continually adapt our strategies to emerging challenges and unforeseen consequences.

 

Tools used so far

GenAI tools I've used so far are


9 views0 comments

コメント

bottom of page