Industry Pioneers Explain How AI Improved Their Efficiency And Accuracy In 2022

Artificial intelligence is now considered indispensable to companies across various industries. Automation not only unclogs the process, but also adds speed and accuracy to everything - of course, if the AI algorithm has been fed the right data sets.

While AI is not perfect yet and is susceptible to making mistakes, a form of arrangement that enables machine learning with assistance from human beings can deliver magical results - be it in building software, diagnosis of diseases in healthcare, performing extensive data-heavy research, and quickly adapting to changes.

Pexels

We caught up with a few pioneers in tech to see how AI made their lives and operations easier in 2022.

The importance of AI

"Artificial Intelligence (AI) is one of the most exciting areas in technology today, and [we have] been leveraging its deep expertise in AI to [help] customers obtain significant insights through data analysis, increase efficiency, scale their businesses, and create better engagements with end-users," said Sanjeev Azad, Vice President - Technology, GlobalLogic, a digital services company.

Sanjeev Azad, Vice President - Technology, GlobalLogic

While enabling AI itself is necessary, data has a key role to play. "While a lot of work is happening in the AI modelling space, the data space is becoming equally important. The kinds of data one uses to build or fine-tune models are becoming critical. One important achievement that we have made is to realign ourselves and bring more attention and focus to our data capabilities around AI," said Anurag Sahay, Managing Director and Head of AI and Data Science, Nagarro, a digital product engineering company.

Also read: Metaverse, AI, And Robots: 5 Tech Trends That Defined The Year 2022

Anurag Sahay, Managing Director and Head of AI and Data Science, Nagarro

"Another key thing we are doing is investing a lot more energy and competency in machine learning operations (MLOps), especially in production and deployment. Because there are so many models that are being deployed and so many ideas that are getting productionised, we feel that having efficient pipelines to lower the company＊s cloud costs is going to be very important. Moreover, we are building a lot more insight platforms rather than predictive platforms, which provide us with a ground for discovering many more use cases for AI," added Sahay.

Pexels

Use value of AI

Artificial intelligence tools shot to fame this year - be it to create dystopian depictions of humanity in centuries from now, or using GPT-3 to weave stories that would have taken hours if done by humans, or to simply create fun avatars of oneself using apps like Lensa.

But within industries, AI has a more useful role to play. For instance, content creation could become more simplified with the use of AI. "[We] created an industry agnostic content generation solution. By leveraging open source GPT2/3 models, we used the concept of transfer learning to create our own custom models which helped in generating new content and repurposing existing content. This solution leverages advanced infrastructure such as GPU, TPU to train complex models," said Azad from GlobalLogic.

Also read: Disney's New AI Can Change Actors' Age Through 'Re-Ageing' Process

Besides this, AI can be used in a variety of settings. For instance, GlobalLogic created an intelligent solution for a "smart bed" maker that would automatically adjust bed firmness while monitoring one's sleep quality in hopes of improving it. That's not all! AI's power can be leveraged in helping people maintain finances more efficiently. "Using machine learning, we were able to accurately make recommendations on how to effectively create and manage short term & long-term financial goals," added Azad.

Artificial intelligence and efficiency

"AI is definitely helping organisations become more efficient. Consider the task of inputting information from an 80-page PDF document with nested hierarchical tables and numerical values into a digital system. If an activity like this would have cost X amount of time and X amount of effort in the past, it now takes only 0.3X of the cost and total time that it used to take. That translates to a 70% improvement in automation and efficiency, which is now very standard. These are numbers from actual use case implementations that we are performing for enterprises," said Nagarro's Anurag Sahay.

Pexels

GlobalLogic's Azad explained how AI and automation may not be used interchangeably, with the former being a subfield of computer science dealing in "development of intelligent systems which can perform tasks that would normally require human intelligence, such as learning, problem-solving, and decision-making."

Automation, he added, "is the use of technology to perform tasks without the need for human intervention. Automation can increase efficiency by reducing the time and effort required to complete a task, as well as reducing the possibility of human error. For example, an automated manufacturing process can produce goods faster and with fewer defects than a manual process."

iStock

It's not a cakewalk! If not implemented correctly, automation can lead to "inefficiencies." "For example, if an automated system is not properly maintained or is not designed to handle specific tasks, it can fail or produce suboptimal results, resulting in waste and decreased efficiency," Azad explained.

"The effectiveness of AI in increasing efficiency depends on numerous factors, including the quality of the AI system, the specific task being automated, and the overall business strategy. In some cases, AI may not be the most effective solution for increasing efficiency, and other approaches, such as process redesign or manual labor, may be more effective."

Also read: AI Tool Shows How Earth's 20 Major Cities Would Look After Climate Changes

While most humans see artificial intelligence for its efficiency, there's a lot more potential waiting to be used. "AI and automation not only improve efficiency but also effectiveness and experience. Furthermore, we should consider AI not only from an efficiency standpoint, but also from the possibility of using AI to solve problems that are currently unsolvable or to create new opportunities for business and the human race. Limiting AI to efficiency will not be fair to what this powerful technology has to offer," said Hasit Trivedi, CTO Digital Technologies & Global Head 每 AI, Tech Mahindra.

Hasit Trivedi, CTO Digital Technologies & Global Head 每 AI, Tech Mahindra

"We have adopted a comprehensive approach to address biases in AI by carefully selecting and cleaning data, testing systems for biases, and implementing processes to mitigate any identified biases. By doing so, we are working to ensure that our AI systems are fair and unbiased and that they can be used to benefit all users," Trivedi added.

Where does India stand?

"In 2022, we navigated the bend to lead the digital transformation and innovation disruption, powered by new-age technologies, and accelerated its amalgamation across operations and workflows. Fueled by India＊s promising growth prospects, 2023 is truly going to be a ＆Techade＊, given the focus on mega trends such as digitalisation, R&D and innovation, local manufacturing, global offshoring, among others. As India is set to lead the G20 agenda in 2023, a lot of focus will be on India＊s capabilities to power sustainable economic growth and development and thereby setting a precedence for others to follow," said Rajiv Bhalla, Managing Director, India & Vice President APAC, at Barco.

Rajiv Bhalla, Managing Director, India & Vice President APAC, at Barco

Also read: Impressive AI Bot 'ChatGPT' Writes Essays, Songs, And More: Here's How To Use It

According to Dr. Yajulu Medury, Vice Chancellor, Mahindra University, AI and automation may be used to improve the scope and delivery of education to students in a future-ready India. Referring to India's new education policy, Medury said that its effective implementation "will aid educational institutions to adopt innovative pedagogical initiatives and build a more holistic multidisciplinary approach towards higher education."

"The resulting multidisciplinary approach to education will help us produce a new generation of individuals capable of working across cutting-edge technology domains that are fast gaining prominence, such as Artificial Intelligence (AI), 3-D machining, big data analysis, machine learning, biotechnology, and nanotechnology," Medury added.

What do you think about the role played by AI in making this year more future-forward? Let us know in the comments below. For more in the world of technology and science, keep reading Indiatimes.com.