AI and Transformation – A memorial talk
Thanks to a recommendation by my friend Dr Gopal Krishna Raju, I was invited to deliver a memorial lecture today at the prestigious Loyola College, Chennai. The event was conducted by their PG & Research Department of Commerce – a one-day seminar in memory of Prof A G Soundara Rajan.
The topic was Artificial Intelligence and (Business) Transformation. I wasn’t interested in playing in my voice a canned presentation or something that was way more of engineering for an audience consisting of Commerce discipline Post-Graduate students from a number of colleges in Chennai.
Instead, I prepared a (custom) talk for them – starting with a brief introduction of AI/ML, proceeding to the possibilities and limitations with the technologies, and ending with a few examples across industries.
In the one-hour talk I had the following covered (in brief, but not in rush):
- Every Industry is becoming a Software one
- What’s Digital Transformation?
- I am not an engineer – Why should I care about AI / ML?
- AI is not new. We can extend the history to the 1308 Catalan poet and theologian Ramon Llull publishing Ars generalis ultima (The Ultimate General Art), his imagination to create new knowledge from combinations of concepts. Modern-day Machine-Learning (ML) can be traced to John McCarthy around 1956.
- Relationship of AI, ML, Neural Network and DNN.
- What’s supervised learning with examples from Andrew NG’s talk (3 Feb 2017)
- What AI cannot do (yet)?
- Examples: Self-Driving Car, X-Ray Diagnosis
- What do businesses need to implement AI successfully – Tech expertise (Or) Domain Knowledge?
- Top Challenge for AI in the Enterprise
- How to win with AI? – Reference: Winning with AI by MIT Sloan
- Steps to do Transformation (in brief) – Reference: Mr Jaspreet Bindra’s “The Tech Whisperer”
- More examples: Machine Learning is happening across companies, Personalization at Amazon, Forecasting at Amazon.com, Object and scene detection, FinTech forms automation, The usage of AI at The Metropolitan Museum of Art, New York (MET) to help people to connect with Art, AI for Environment & AI for Agriculture
- Then I showed a coding example using OpenAI’s GPT2 to generate a new review after the system has learned from the hundreds of past movie reviews in my blog.
- I close the talk with a fun quiz to deduct Fake or Real from ThisPersonDoesNotExist generated images and the example of Mark Zuckerberg face being replaced with Star Trek Commander Data using DeepFakes’ Faceswap technology.