Welcome to our Agile AI Insights newsletter — where AI Innovation Meets Agile Execution

What Agile AI Insights will do for you:

• Provide the latest developments and news in AI & Generative AI for business

• Showcase the latest AI training and career opportunities

• Discuss AI project challenges and how to manage these projects successfully as well as deep dive into AI (ML & Generative AI) use cases and applications

  
  

Generative AI for Business News You May Have Missed

• Accenture invests $3 billion into artificial intelligence

  Accenture announced it will invest $3 billion over three years into its data and AI practice. (read more)

  
  

• Visa launches AI consulting practice:

  Payments giant Visa announced the formal launch of a global artificial intelligence (AI) advisory practice. (read more)

   

• IBM Consulting invests heavily in generative AI on AWS platform

  IBM Consulting has expanded its relationship with Amazon Web Services (AWS) to help more mutual clients derive value from generative artificial intelligence. (read more)

  
  

• PwC US to invest $1 billion in generative artificial intelligence

  PwC invests $1 billion over the next three years to scale its generative AI capabilities. (read more)

  
  

AI Training & Certifications

• Andrew Ng Founder of DeepLearning launches “AI for Everyone” course:

  AI for Everyone”, a non-technical course, will help you understand AI technologies and spot opportunities to apply AI to problems in your own organization.

   

• New 1h course by WhyLabs on “Quality and Safety for LLM Applications” course:

  This class shows you how to mitigate hallucinations, data leakage, and jailbreaks. Incorporating these ideas into your development process will make your apps safer and higher quality.

  
  

• Introduction to Artificial Intelligence (AI) by IBM on Coursera

  In this course you will learn what Artificial Intelligence (AI) is, explore use cases and applications of AI, understand AI concepts and terms like machine learning, deep learning and neural networks.

  
  

An introduction into Agile Delivery and the fundamentals of types of AI & Generative AI

Welcome to our first newsletter. We are happy you are here.

Everyone is talking about Artificial Intelligence or AI, and using acronyms like ML, LLM and Generative AI.

But what do they all mean and how do they come together?

In future newsletters we will dive deeper into AI, but I thought it would be good to first demystify common AI acronyms and the difference between AI, ML, LLM, and Generative AI.

Fundamentals of types of AI & Generative AI

AI is the theory and development of computer systems able to perform tasks normally requiring human intelligence.

Machine Learning (ML), which falls under the AI umbrella, is about training computers to learn from data. Unlike traditional programming where we dictate every step, in ML, algorithms sift through data, detect patterns, and make informed predictions on new, unseen data. It's a shift from explicit coding to data-driven learning. Machine learning is the technology in many of the applications and services we utilize today like the recommendation systems in Youtube, Netflix or Spotify.

Neural Networks (NNs), also known as artificial or simulated neural networks, are a key part of machine learning, modeled after the human brain. They're made up of layers of artificial neurons: an input layer, several hidden layers for processing, and an output layer for results. An ML Model makes decisions based on what it has learned from the data, whereas a neural network arranges algorithms so that it can make decisions reliably on its own.

NNs are crucial in deep learning, where they work behind the scenes to analyze and interpret complex data.

Deep Learning (DL) is a type of Neural Network (NN) with a bit more complexity. The 'deep' in Deep Learning refers to the many layers it has – think of a neural network but with over three hidden layers. This depth makes DL great for sorting out complicated problems in the real world. You've actually seen deep learning in action in everyday tech, like the facial recognition or automated driving in vehicles like Tesla.

Generative AI, or GenAI, is a branch of Deep Learning. It's a smart AI tech that can whip up various kinds of content like text, images, audio, and video, all by learning from existing content.

Then there are Large Language Models (LLMs), a specific type of GenAI. LLMs are all about creating text that sounds like it was written by a human. They get really good at this by studying heaps of textual data. It's important to note that LLMs are a niche within machine learning, focused on natural language processing. Meanwhile, Machine Learning (ML) is a broader field with a whole array of algorithms and methods for different purposes, like making predictions or decisions based on data. A prime example of LLM technology in action? ChatGPT – it's pretty well-known for using this tech.

If we look at Machine Learning from the prospective of how learning is achieved by the models, we have three main styles: Supervised, Unsupervised, and Reinforcement Learning.

In Supervised Learning, think of the computer like a student learning from a textbook with answers. It's shown examples (input) and the right answers (output). The training data needs to be well-organized and labeled for this to work. Supervised Learning is used for two main tasks: Regression, where it predicts a continuous value like a house price, and Classification, like figuring out what's in a picture. Supervised Learning is pretty well established and there are plenty of pre-made algorithms, but the catch is you need clean, labeled data.

Unsupervised Learning is more like a detective, finding patterns without knowing the answers upfront. It’s used for grouping similar things together (Clustering) or spotting the odd ones out (Anomaly Detection). The cool part? No need for labeled data or deep domain knowledge. But, it's trickier to get right and needs extra analysis to make sure the findings make sense.

Reinforcement Learning is all about trial and error to maximize rewards, sort of like learning to play a game by practicing. It's great for stuff like teaching computers to play Chess or Go. The computer learns which moves bring wins and which don’t. These algorithms can get really specific results, but they work best when there's a clear reward and the ability to simulate scenarios. They need lots of practice runs to really get the hang of things.

Consider this an introduction to the fundamentals of Artificial Intelligence.

In conclusion, the landscape of Artificial Intelligence (AI) is vast and dynamic, encompassing various technologies from Machine Learning (ML) to Generative AI (GenAI), each playing a unique role in advancing how computers imitate human intelligence. ML stands at the forefront of this revolution, transforming the way machines learn from data. Neural Networks (NNs) and Deep Learning (DL) add layers of complexity and depth, enabling machines to tackle more intricate tasks, such as facial recognition and automated driving. Generative AI, particularly through Large Language Models like ChatGPT, pushes the boundaries further by creating content that mirrors human creativity. The diverse learning styles - Supervised, Unsupervised, and Reinforcement Learning - each contribute distinctively to AI's ability to adapt, predict, and evolve. As we continue to explore AI's potential, its applications in business and everyday life are bound to expand, offering exciting prospects for the future.

In future sessions we will dive deeper in these topics and look at the applications in business and their real-life examples.