Recently, for an episode of Simulated Reality, a video podcast series by Analytics India Magazine, we had Elad Ziklik, VP, Product Management for AI Services and Data Science at Oracle, in the house.
Elad is a Computer Science graduate from the University of Haifa and has over 21 years of experience. He has earlier worked with Intel as its Lead- software team and group, post which he spent 14 years in Microsoft where his last position was that of Partner Director of PM- Knowledge Mining. Elad joined Oracle in Seattle in September last year. He believes in integrating the power of artificial intelligence (AI) and machine learning (ML) to create unique experiences and products as an experienced and enthusiastic product leader.
During the session, Elad delved deep into some critical aspects of artificial intelligence and how organisations can leverage its full potential, the challenges involved, and Oracle’s key AI capabilities and strategies.
Subscribe to our Newsletter
Join our editors every weekday evening as they steer you through the most significant news of the day, introduce you to fresh perspectives, and provide unexpected moments of joy
How we need to approach AI differently v/s regular software
During the conversation, Elad emphasised that AI is a paradigm shift, and can’t be approached the same way as regular software, given most people are used to the binary world of software – where every problem seems solvable, with time, resources and the right engineers.
“Working with AI and ML is different from regular software,” Elad said. He added that in AI, one needs to experiment and see what works, much like in research. While most leading cloud providers are busy building general-purpose platforms and models, they are missing out on building models that work for the actual business problems of the last mile. “These companies are building amazing technologies, models, and the GPTs with billions of parameters. But how customized are they for your business/enterprise? For example, a general-purpose platform will still not know how to spell my name. Until we solve for that level of simplicity and ease, AI will keep falling short in real-world scenarios,” he explained.
Providing a solution to this challenge, Elad emphasises that while technology providers try building the biggest and most impressive models of all time, what he wants to do is “get away from trying to make AI cool”, and instead “make AI boring”, so much so that it’s ubiquitous – it’s always there in the background and you wouldn’t even notice it as a part of your life or business. “I call it ‘invisible AI’, just like electricity – you flip a switch and the light comes on; you don’t need to worry about the wiring behind the walls etc, right? That’s how we need to make AI work”, adds Elad.
“If AI needs to be successful, technology providers need to make it very easy. They need to make it such that there is no need for special unicorns, data scientists, geniuses that can go and do super sophisticated stuff. It needs to be everyday AI, something that works for the boring business problems – very vertical/domain-specific and address the last mile need. That’s what we’re focused on at Oracle”, Elad said.
Oracle’s Key AI Use Cases
Elad further provided some offbeat examples of how Oracle’s customers benefit from the company’s AI services. He said that in an Indian context, Oracle has been using computer vision technologies to solve many problems. For example, during the COVID-19 crisis, pandemic-induced panic made many rush to their nearest ATMs to stock up money. Banking and financial institutions had to ensure that they stocked cash in the right ATMs and at the right time to ensure seamless services to their customers.
Secondly, another company uses AI to clean waters in rivers and lakes. It uses a robotic machine that floats in the river water, picking up debris and ensuring that the water body remains clean. However, the more challenging issue was figuring out where the debris or waste was. Drone technology is being used to find out the spots and identify the garbage using computer vision. Once the spot is figured, the cleaning bots reach the spot with the help of GPS and clean it.
“I find these kinds of autonomous devices very interesting and we will probably see a lot more of them in the near future,” Elad added.
Analytics India Magazine asked Elad about the trends he foresees regarding AI in 2022 and beyond. To that, the VP- Product Management for AI Services and Data Science said that he has been noticing a rise in large models that are being trained on billions of parameters. Secondly, he notes that people are starting to move away from big data into ‘little data’. The coming together of these two have already started, and with that, AI models have become ubiquitous.
Additionally, Elad predicts there will be a rise of ethics in AI, with more regulations and expectations coming into space. He is also of the opinion that we shouldn’t be surprised if companies create the role of ‘Chief AI Officers’ or something similar in the near future, given how important AI has become in today’s world.
During the video podcast, Elad reveals Oracle’s AI strategy and how its OCI-AI services can help enterprise customers. He focuses on how Oracle’s solutions are domain-specific and aimed at creating a consistent experience, making it easier for developers to tune the last mile of AI accuracy to their specific domain, industry and business. “We are making AI that solves all your business scenarios– understanding text with natural language processing and computer vision in speech. We have also introduced a set of decision services. These are business-supported services that help make simple or sophisticated business decisions– like anomaly detection, focusing on demand and resources, how to personalise and recommend offers to users, and how to optimise the next action of a business’s journey”, he added.
Oracle is trying to make the process of retraining and customisation much easier for the developers.
Before concluding the conversation, Elad was asked to describe AI in one word or phrase. To which the VP- Product Management for AI Services and Data Science said, “AI is just getting started,” while people might believe that AI has been there for decades and companies might claim to be leaders in the space, Elad believes that there is no clear leader in the space yet and we are just getting started.