On August 18, 2023, the Business Intelligence Group Committee (BIG) at Scrubbed spearheaded the groundbreaking AI Hybrid Seminar. The event was attended by more than 600 Scrubbees both in person at the Scrubbed Tech Center and virtually via Zoom. The seminar aimed to empower Scrubbees for Adaptive Innovation. The BIG Committee supports and guides all employees in embracing Business Intelligence as an integral part of their client service. Seminars like this one complement their mission to introduce, promote, and integrate Business Intelligence capabilities as a core strength across all service teams.
The Scrubbees enthusiastically participated in the engaging icebreaker – The BIG Quiz! This quiz tested our general knowledge of AI and served as a great warm-up before the Introduction to AI presentation. The presentation was led by Daryl Peralta, Lead Engineer at Samsung R&D Institute Philippines, also a part-time lecturer at the University of the Philippines Diliman. He holds a Bachelor’s degree in Electronics and Communications Engineering with magna cum laude honors, and a Master’s degree in Electrical Engineering from the same institution, specializing in machine learning and computer vision.
Demystifying AI: Key Takeaways from the Seminar
• Artificial Intelligence (AI) can be Narrow AI (e.g. Gmail’s Smart Compose or Google Translate), which performs specific tasks, or General AI, also called Artificial General Intelligence (AGI), which aims for full autonomy and self-awareness. But, this concept remains a vision up to the present day. • Advancements today are driven by Machine Learning, the ability to learn without explicit programming, with a subset known as Deep Learning, from which Facial Recognition has emerged. • The AI concept, specifically the Neural Network, was proposed in 1957, but it has only recently gained significant attention due to the following factors:-
-
-
- The abundance of BIG DATA (e.g., Wikipedia, YouTube, Kaggle)
- The availability of powerful Hardware (e.g., Colab, GEFORCE RTX)
- The development of software frameworks aimed at accelerating research (e.g., PyTorch, TensorFlow)
-
-
• In the realms of finance and business, Machine Learning removes the need for entering formulas and highly specific instructions into Excel. Instead, it utilizes data to enable machines to discover these instructions.
• As famously stated by Andrej Karparthy (Director, OpenAI), “The hottest programming language is English.”
• AI proves highly useful, yet it possesses limitations. Currently, Large Language Models (LLM) like GPT-3, which are trained to understand human-like language can only deduce rules when fed with data, and the complexity of the underlying reality of the language makes it challenging to capture.
• Large-scale deep learning’s environmental sustainability is questionable due to its carbon footprint from training. However, ongoing efforts aim to make AI more environmentally friendly and even utilize it to help solve climate and sustainability-related problems.
• Algorithmic racism is a potential concern in AI systems resulting from bias from data inputs and due to their black-box nature, preventing them from explaining the reasons behind their outputs.
• While tools are available to identify content plagiarism or AI-generated articles, this gap is likely to narrow as AI continues to advance.
• When dealing with Scrubbed or any confidential data, it’s essential to review the terms and conditions of the product (e.g., ChatGPT). Therefore, entering sensitive and confidential information is not recommended, as control over data destinations may be limited.
Embracing change and driving innovation is one of Scrubbed’s aspirations and everyone’s responsibility. We don’t need to be AI experts and engineers in order to come up with innovative ideas and out-of-the-box solutions, as emphasized by Diana Peralta, BIG Chairman at Scrubbed.
Exploring AI introduces new avenues for generating innovative solutions, enabling us to offer more efficient and higher-quality service to clients in today’s rapidly evolving business landscape. However, it’s crucial to acknowledge its limitations and potential challenges, prompting us to use AI responsibly.
AI Learning Resources
- Online Courses:
- Deep Learning Book
- YouTube: www.youtube.com/@YannicKilcher
- Learn Computer Programming, Algorithms, and Data Structures.
- Join Kaggle competitions
- Read research papers
