AWS ML engineering manager evaluates the social impact of AI

Ankit Sirmorya, Amazon’s machine learning (ML) engineering manager, is an emerging thought leader in the artificial intelligence (AI) and ML space. He serves as the community lead for Global AI Hub, a group dedicated to AI that is growing at a rapid pace. He reviews software engineering and data science-related books and provides feedback after in-depth analysis. Sirmorya also has a YouTube channel, Tech Takshila, which has a worldwide audience. 

Sirmorya was awarded a spot on the 40 Under 40 Data Scientists list by Analytics India Magazine in 2022. He was also a finalist for the U.K.’s Computing AI Professional of the Year in 2022.

In an exclusive Q&A interview with VentureBeat, Sirmorya discusses how AI is changing the way we interact with technology, how it is affecting our everyday lives, and the promise it holds for the future. In addition, we discuss his journey from India to become a machine learning expert in one of the world’s top tech companies.

VentureBeat: What developed your interest in machine learning?

Sirmorya: A childhood interest in programming led me to pursue a Bachelor of Science in Computer Science. While studying for my undergraduate degree, I had a few courses related to machine learning that sparked my interest. Following that experience, I pursued a graduate degree in machine learning at the University of Florida. I completed several data science courses and conducted research in renowned professors’ labs. My job at Amazon involved applying machine learning to customer experience problems. Moreover, Amazon’s Machine Learning University has allowed me to gain a deeper understanding of data science and grow my interest in it.

VB: How do consumers interact with AI in their daily lives that they don’t realize?

Sirmorya: Many people do not even realize how often they interact with AI. For example, AI provides video recommendations on streaming platforms such as Netflix and Amazon Prime Video. It also generates news feeds on social media platforms such as Instagram and interacts with voice assistants such as Google Home and Amazon Alexa. The use of facial identification is another application of AI. At Amazon, I work on Alexa shopping, where my team develops machine learning-based solutions that send personalized hints and notifications to customers.

VB: How do you define the practical business differences between AI and ML?

Sirmorya: When it comes to big data, analytics, and the broader waves of technological change sweeping the globe, both terms tend to appear frequently. Artificial intelligence refers to machines capable of performing tasks in a way that would be considered “smart.” Machine learning, on the other hand, is the current application of AI centered on enabling machines to learn for themselves by accessing data.

Artificial intelligence can be divided into two basic types: applied and general. Applied AI is far more common; systems that trade stocks and shares intelligently or navigate autonomous vehicles fall under this category. Generalized AI – a system or device that in theory can handle any task – is less common, but the most exciting developments are taking place today in this area. In addition, it has contributed to the development of machine learning. ML should be thought of as the current state-of-the-art (rather than as a subset) of artificial intelligence.

A machine learning application can read text and determine whether the author is complaining about or congratulating someone. Additionally, they can listen to music to determine whether it will make someone happy or sad and find other pieces that match the mood (such as Spotify). Some can even compose their own music expressing the same themes, knowing that their fans will enjoy it.

VB: What are some significant current use cases for AI and ML?

Sirmorya: A successful AI company understands how AI technology is consumed. Artificial intelligence is an enabling technology that can be applied to solve a variety of business and industrial problems. In the 1980s, SQL databases allowed structured data to be stored and retrieved in tabular form. As a result, billions of dollars were generated in the enterprise resource planning and customer relationship management industries. AI will also enable a wide range of new applications. Other examples:

• Vision algorithms can detect manufacturing defects, identify shoplifters, and navigate autonomous vehicles, for example. 
• We can apply natural language processing for many purposes, such as reviewing customer sentiment on social media, reviewing legal documents for contract completion, evaluating resumes and CVs, and determining financial trading opportunities. 
• Medical transcription algorithms also use speech recognition and transcription. 
• As an enabler, AI can be used by corporations and other organizations to solve problems. 

VB: What types of future use cases might we see that will deploy AI?

Sirmorya:  Artificial intelligence and data analytics are poised to revolutionize many sectors in the near future. There have already been significant deployments of AI in finance, national security, health care, criminal justice, transportation, and smart cities that have altered decision-making, business models, risk mitigation, and system performance. There are substantial economic and social benefits resulting from these developments. Things that were once considered impossible are now becoming possible thanks to artificial intelligence, such as driverless cars. Fast GPUs and access to training data are both key enablers of driverless cars. 

AI is poised to revolutionize familiar activities such as dining. Restaurants could use AI to determine which music to play based on the interests of their guests. It is even possible for artificial intelligence to modify the appearance of wallpaper based on what it anticipates the aesthetic preferences of the crowd will be.

AI is even expected to free digital technology from its two-dimensional, screen-imprisoned form. It is possible that the primary user interface will be the physical environment around us. In order to play a game, interact with a web page, or read an e-book, we have always used a two-dimensional display. With artificial intelligence and the internet of things, the display won’t be the main interface anymore — it’ll be the environment. Whether it’s connected buildings or connected boardrooms, people will design experiences around them. There will be 3D experiences that you can actually feel.

VB: How does AWS offer AI services to help solve problems?

Sirmorya: Pre-trained AWS AI Services provide ready-made intelligence for customer applications and workflows. AI services can be easily integrated into customer applications to address common use cases, such as personalized recommendations, modernizing a contact center, improving safety and security, and increasing engagement. Customers get quality and accuracy from continuously learning APIs because AWS AI services offer the same deep learning technology that powers Amazon.com. The best part is that AI Services on AWS do not require any experience with machine learning.