Transformers

Transformers are considered revolutionary in the field of AI since this deep learning architecture was first written in Google’s research paper “Attention Is All You Need” by Vaswani et al., in 2017.  The transformer model has now become the foundation for many state-of-the-art AI models and products like ChatGPT, Google Bard, Midjourney…

Before the transformer model, recurrent neural networks (RNNs) where data is processed sequentially, was a popular language AI architecture.  However, the transformer architecture uses the attention mechanism to look at the relationships among all the words simultaneously from many different angles and determine which words carry more weights.

Let’s ask GPT-4 to explain transformers in an easy-to-understand way.

Here's the intermediate explanation of transformers from GPT-4.

LSTM

 

Long short-term memory (LSTM) network is an improved version of recurrent neural network (RNN) where the system excels at sequence prediction tasks and capture long-term dependencies.  LSTM was first published in a technical paper by Sepp Hochreiter and Jürgen Schmidhuber.

Let's ask GPT-4 to explain LSTM at a basic level.

Here's the intermediate explanation of LSTM from GPT-4. 

LSTM network is effective at capturing long-range dependencies and retaining essential information over longer sequences.  However, LSTM network architecture is still a sequential model where the sentences are processed word by word.  In the next post we will explore an improved network model that revolutionized sequential data processing.

Mindmap Generator

 

In the previous mindmap of ServiceNow’s Governance, Risk, and Compliance (GRC) product created with GPT-4 + AI Diagrams, the map was not the optimal outcome that I envisioned.  Since I like what MyMap.AI's Mindmap Generator can do, I’ve requested it to produce a mindmap for ServiceNow's GRC.  

The result was a professional-looking mindmap with comprehensive product information.  This mindmap also has editing features that allow users to add or remove nodes and texts to the users' specifications.  Click on image below for graph with legible texts.

In the age of AI, the pace of technological advancement and innovation is only accelerating.  The competitive advantage is going to shift to people who know how to access and summarize a large amount of information, present the result in a visually engaging way, and do all of this quickly.  Work is going to be an exhilarating ride from now on with AI.

MyMap.AI

 

There has been a proliferation of AI-powered applications on the market and it can be an overwhelming experience for users due to ever-increasing selections.  I’ve recently come across MyMap.AI which helps you to create, organize, and share your ideas.  MyMap.AI also has a set of free tools, one of which is the Mindmap Generator which generates mindmap. 

Let’s test it out.  I’ve requested MyMap Mindmap Generator to create a mindmap for a Go-to-Market strategy for ChatGPT.  I was impressed that the mindmap presented a good strategy with a comprehensive action list.  Even more remarkable was the mindmap created in only 15 seconds, in what traditionally would take hours to produce.  Click on image below for graph with legible texts.

The power of AI is amazing as it helps people to become more creative and productive.  I believe AI will enhance the quality of work going forward (providing there's safeguard in place to prevent AI from going rogue).  AI is here to stay so people should accept and adapt to this new reality the sooner the better.

Season's Greetings

 

It’s the holiday season so I would like to send a message of hope to everyone.  GPT-4, Claude, and Bard were prompted to create a haiku about hope and world peace.

GPT-4 produced the best haiku because it’s more beautiful and poetic:

Dawn’s light breaks the night,

Seeds of hope in bloom, so bright,

Peace on earth takes flight.

Best wishes and good luck in 2024!

RLHF

 

Reinforcement Learning from Human Feedback (RLHF) has been employed to train and fine-tune large language models to achieve better performance for artificial general intelligence (AGI).  Here are the explanations of RLHF (1) one from ChatGPT at simple level, and (2) the other from Brian Roemmele (@BrianRoemmele) that is concise and descriptive.

1)  From ChatGPT:

2)  From @BrianRoemmele:

RLHF stands for "Reinforcement Learning from Human Feedback."  It's a technique used in machine learning where a model, typically an AI, learns from feedback given by humans rather than solely relying on predefined datasets. 

This method allows the AI to adapt to more complex, nuanced tasks that are difficult to encapsulate with traditional training data.

In RLHF AI initially learns from a standard dataset and then its performance is iteratively improved based on human feedbacks. 

The feedback can come in various forms, such as corrections, rankings of different outputs, or direct instructions.  The AI uses this feedback to adjust its algorithms and improve its responses or actions. 

This approach is particularly useful in domains where defining explicit rules or providing exhaustive examples is challenging, such as natural language processing, complex decision-making tasks, or creative endeavors.

ServiceNow GRC

 

Environment:  ChatGPT Plus, GPT-4.

Objective:  Explain complex topics with summary and mindmap.

ServiceNow has recently been named as a leader in Governance, Risk, and Compliance (GRC).  GRC, one of 36 workflow automation products of ServiceNow, is backed by AI innovation throughout the risk management lifecycle.  The product helps corporations to connect business, security, and IT, and to handle risk with confidence.

Using the combination of GPT-4 + VoxScript + AI Diagrams, I prompted GPT-4 to explain ServiceNow GRC from the perspective of the CEO Bill McDermott.  The result described the product’s features with sophisticated business language.

Then GPT-4 was prompted to create a mindmap visualizing ServiceNow GRC product.  This mindmap organizes the key aspects and provides excellent summary of the intelligent product.

Verdict:  The quality of the responses from GPT-4 depends greatly on the way the prompts were instructed.  Prompt wisely to reap the benefits from GPT-4. 

On a different note, ServiceNow has a family of 36 AI-powered workflow automation products that can help your company on the digital transformation path.  You can peruse the company website to determine which products are beneficial to your company's needs.