On a rainy Monday, over 50 business leaders congregated early in the morning in Dusit Thani, Makati to listen to Dr. Mohanbir Sawhney talk about the applications and implications of Artificial Intelligence for Business Leaders. In recent months, AI has gained ground in the global zeitgeist, with the recent improvements in computing power and the rise of LLMs life ChatGPT. Now, more than ever, Artificial Intelligence is carving out a large chunk of our daily lives and the business world is being disrupted at a pace faster than the previous tech boom.
Dr. Mohanbir Sawhney is a faculty member at the Kellogg School of Management, which is part of Northwestern University in Evanston, Illinois, USA. He holds the position of the McCormick Foundation Chair of Technology at Kellogg. Throughout his career, he has taught in various executive education programs and has contributed significantly to the growth of the technology and marketing curriculum at the school.
Introducing AI and Generative AI
The first session of the workshop, entitled An Introduction to AI and Generative AI, began with a reflection on learning agility, where it was highlighted that just six months ago, no one knew anything about ChatGPT; now, it has become a popular program at Kellogg for executives. From a concept written on a paper napkin, this has been executed into a program that will be fully subscribed for in the next 3 semesters.
To begin one’s journey into AI, Dr. Sawhney asked the audience to classify themselves among 3 levels: A. This is all quite new to me, B. I have been involved in some AI projects, and C. I have been responsible for driving initiatives within my organizations.
Level 1: Beginner – This is all quite new to me.
Dr. Sawhney suggested beginners start by conducting research and learning about AI technologies, their applications, and potential benefits for your industry and business. Consider engaging with AI consultants or experts to gain insights into AI’s capabilities and how it can be applied in your organization. Identify areas in your business that could potentially benefit from AI, such as customer service, marketing, or supply chain management. Begin by piloting small AI projects to gain hands-on experience and understanding of their potential impact on your business.
Level 2: Novice – I have been involved in some projects.
For those with some level of AI knowledge, assess the success and impact of previous AI projects within your organization. Collaborate with AI experts and internal stakeholders to identify new opportunities for AI implementation and expansion into other business areas. Develop a business case for AI initiatives by analyzing potential costs, benefits, and ROI. Consider both tangible and intangible benefits. Implement AI-powered customer experience solutions for front-office functions, such as personalized marketing and customer service interactions. Look into AI applications for back-office functions, like optimizing HR processes or improving supply chain efficiency.
Level 3: Expert – I have been responsible for driving initiatives within my organization.
For those with clear expertise in Artificial Intelligence applications, Dr. Sawhney advised this group to lead cross-functional teams to explore and implement AI initiatives across various business areas. Define clear goals and KPIs for AI projects to measure their success and impact on the organization’s overall strategy. Integrate AI into your company’s long-term business strategy to drive competitive advantage and innovation. Investigate the use of AI in operations, such as using predictive analytics for supply chain optimization or AI-powered quality control in manufacturing. Stay abreast of the latest developments and research in AI to anticipate future challenges and opportunities.
Examples of AI Apps
Dr. Sawhney then went on to show the various applications that users have been dabbling in. He showcased “Botto.com,” an AI-driven platform that produces paintings. The generated artwork is then sold as NFTs (Non-Fungible Tokens) using cryptocurrencies like Ethereum and Bitcoin. He also mentioned “This Person Does Not Exist,” a website that utilizes Generative Adversarial Networks (GANs) to create random images of non-existent human faces. GANs consist of two networks—one generates content, while the other discriminates real from fake (also used in deep fakes). In addition, Dr. Sawhney highlighted AI-generated videos that eliminate the need for hiring models. Companies can specify race, size, and other attributes, and AI dynamically matches them with suitable models. This approach saves resources, as there is no requirement for extensive photoshoots or model hiring.
AI in Education
Being an educator, Dr. Sawhney was excited about the implications of AI in education and the profound effect it will have for the careers of professors all over the world. To show this, he recorded a 5-minute video on a simple topic. After processing for 10 minutes, the Artificial Intelligence produced a complete course, including 25 quiz questions in a “drag the word” exercise format. Additionally, it generated a summary and transcript of the video content. The AI can also record new videos and create customized learning pathways. The company responsible for this AI technology was established only three months ago, and the professor just learned about it last week. Imagine then, the vast library that can be created by educators all over the world on varying subjects and levels of difficulty and how fast this can be deployed to the marginalized sector.
Origins of AI
Dr. Sawhney then took a deep dive in the history of Artificial Intelligence and how it had come to be in its current state that we know. In its inception in the 1960s, AI used to be knowledge-based, where programs were built on a rules database. Algorithms selected rules to respond accordingly. He showed an old case– a database of 4,000 research papers was used to create an interface for a deodorant challenger in a growing market. Based on the 4,000 documents, the AI was able to generate a strategy given a set of parameters based on geography, gender, economic standing, and other relevant data points. However, the problem with this approach is that it was not smart and was limited to its own inputs. It required constant feeding of rules, making it less adaptable and unable to learn on its own.
1970s: The AI Winter
In the 1970s, there was a lack of computational power which meant that the computational resources available during that era were far more limited compared to today. AI algorithms required significant computing power to process and analyze large datasets, making it challenging to achieve significant breakthroughs. The prevailing AI approach during this period, called “symbolic AI,” relied heavily on rules and logical reasoning. While it showed promise in certain areas, it struggled with real-world complexities and lacked the ability to learn from large datasets. These challenges collectively led to a decrease in AI research funding, the closure of AI laboratories, and a general decline in interest in the field. This period of stagnation in the 1970s and early 1980s became known as the AI winter.
1990s: Machine Learning
The 1990s saw a jump in computational power with the abundance of personal computing and enterprise grade computers. AI was back with a more sophisticated approach to problem-solving called machine learning. Machine learning, a crucial aspect of artificial intelligence, empowers computers to learn from data and experiences without explicit programming. It involves training algorithms with examples to recognize patterns and relationships, enabling them to make informed decisions on new, unseen data. Machine learning encompasses various types, such as supervised learning with labeled data, unsupervised learning for finding hidden structures, and reinforcement learning through interactions with an environment. Common algorithms like decision trees, neural networks, and k-nearest neighbors are employed for diverse applications such as image recognition, natural language processing, autonomous vehicles, and fraud detection. The process involves model training, where algorithms adjust their parameters to minimize errors, and evaluation using separate datasets to ensure effective performance and generalization.
Present: Generative AI
From there, it was a short jump to Generative AI, which is a type of artificial intelligence that utilizes deep learning techniques but focuses on creating new and original content, such as text, video, images, or even code. It produces unique outputs based on the patterns and information it learned from the training data, resulting in content that has not been seen before. One measure of its effectiveness is the Turing Test, where, if a person interacting with the AI cannot tell within 30 minutes whether they are talking to a human or a computer, the AI is considered to have passed the test as a legitimate form of AI.
AI at an Inflection Point
Dr. Sawhney then tackled the reasons why Artificial Intelligence is currently at an inflection point. First, the explosion of data. The amount of data available has skyrocketed, enabling AI systems to make more accurate and informed decisions. This abundance of data has transformed industries like risk underwriting, where previously limited data made assessments challenging. Now, AI algorithms can factor in vast amounts of data to make better-informed risk predictions and decisions. For example, language models like GPT (Generative Pre-trained Transformer) have access to extensive human-written data, allowing them to generate text that seems human-like.
Second, the rise of cloud computing. The availability of powerful cloud computing resources has revolutionized AI capabilities. Cloud platforms like Azure Machine offer access to immense computational power, such as supercomputers with billions of parameters. These resources enable complex AI models to process data efficiently, resulting in more accurate and sophisticated outcomes. Additionally, the cost-effectiveness of cloud computing has made AI accessible to a broader range of users and businesses, with queries costing mere cents.
Third, algorithmic breakthrough. Continuous advancements in AI algorithms have led to breakthroughs in various domains. These improvements have made AI models more efficient, accurate, and capable of handling complex tasks. Researchers and developers are continually discovering new techniques and methods, driving AI’s rapid progress.
Lastly, the availability of investment and funding. AI is attracting significant investment from both big tech companies and startup ventures. The availability of capital has fueled research and development in the AI field. Hundreds of generative AI startups have received funding, contributing to the growth and innovation in the industry.
In conclusion, the first session on Introduction to AI and Generative AI with Dr. Mohanbir Sawhney was a captivating journey into the world of artificial intelligence and its immense impact on businesses and daily life. As AI gains ground in the global zeitgeist, the advancements in computing power and the rise of LLMs like ChatGPT have propelled AI to disrupt industries at an unprecedented pace. Dr. Sawhney’s expertise shed light on how businesses can embrace AI based on their level of familiarity, from beginners learning about AI technologies to experts driving AI initiatives within their organizations.
The journey through the history of AI, from its knowledge-based inception to its resurgence with machine learning, provided valuable insights into the evolution and promise of AI. The current inflection point of AI is attributed to its ability to process massive amounts of data, fueled by cloud computing and breakthrough algorithms, all while attracting substantial investment and funding.
Talk to us to know more about our leadership programs, or send us an email at info@johnclements.com.
