Real Life Applications of Research Areas

Artificial intelligence (AI) and machine learning (ML) are evolving from breakthroughs to mainstream technologies at a fast pace- and in many areas. So grab this wave of opportunity.

At first, it was like a meteor –coming from far away somewhere and sparking off a lot of curiosity. Then it became like water – slowly pouring into many aspects of business and everyday life. And pretty soon, it will be like air- all around us, essential and still invisible.

The sharp rise, and consistent maturity, of AI, are truly remarkable. AI and its subsets like machine learning and deep learning have permeated almost every industry and application today. Just pause and check around your room or inbox- I am sure that AI is now driving a significant chunk of your day.

From a handshake to a bear hug

AI, put simply, is the ability of a computer or a robot or any form of technology to perform task/s that are usually done by humans – and done through simulating human capabilities and human intelligence. For example, machine learning is a specific type of AI program which enables a system to learn and improve from experience without any explicit programming automatically.

Let’s take a quick walk through some reports to get a snapshot of how fast AI has started redefining our work and lives. 

As per a recent O’Reilly annual survey on AI adoption, 26 percent of organizations had AI projects in production, 43 percent said they are evaluating it, and 31 percent are not using it. 

Stanford’s AI Index Report 2022 shows that private investment almost doubled between 2020 and 2021- touching around $93.5 billion. In addition, AI is becoming easy to deploy and quite affordable as time goes by. In 2018, the cost to train an image classification system dipped by 63.6 percent, while training times have jumped by 94.4 percent. Lower training cost but faster training time is a pattern that has echoed in other MLPerf task categories such as recommendation, object detection, and language processing as well.

As per 451 Research’s ‘Voice of the Enterprise: AI & Machine Learning, Use Cases 2021’ survey, 95 percent of respondents say AI/ML was very or somewhat important to their digital transformation efforts. Almost three in four enterprises have either invested in MLOps tools or plan to do so in the next 12 months. It is expected that the adoption of MLOps will go on unabated in 2022. 

According to S&P Global Market Intelligence, AI and its machine learning subset are adopted across enterprises to optimize, automate and augment digital transformation strategies. The analysts here expect take-up of more advanced use cases and the long-standing use cases, such as fraud detection in financial services and churn analysis in telecom. It is also signified that more use-cases will emerge for AI for employee safety in manufacturing, clinical trial analysis in healthcare/life sciences, and vision analytics for infrastructure inspection in telecom.

We have already started feeling AI as a standard ingredient in our lives when we use chatbots in banking, or AI-apps for food ordering, or AI-based algorithms to decide what to watch on OTT. However, many more applications will emerge and gain a stronghold as AI gets more mature, easy to execute, and ROI-friendly.

All these applications are translating into good numbers. The global artificial intelligence services market is slated to rise from $8.24 billion in 2021 to $12.39 billion in 2022, as per data from ReportLinker. It has been explained that a lot of this growth is due to companies resuming their operations and adapting to the new normal while recovering from the COVID-19 impact. The market is expected to reach $58.93 billion in 2026. A significant portion of the artificial intelligence (AI) services market would entail sales of AI services in telecommunications, government, retail, defense, and healthcare. We will also see the strengthening of ‘AI as a service’ as businesses start to use AI for different purposes, without significant initial investment, and with lower risk. However, fuelling this growth would need a good strength on the AI workforce side.

More AI means need for more human capabilities

As seen in the O’Reilly survey, respondents with AI in production, and respondents who were evaluating AI, said that the most significant bottlenecks were lack of skilled people and lack of data or data quality issues (both at 20 percent), apart from problems like difficulty in finding appropriate use cases (16 percent). In addition, the lack of people with relevant cp was shared as a critical challenge by organizations with AI in production and those evaluating AI (20 to 25 percent).

The next set of risks was in unexpected outcomes (68 percent), model interpretability, and model degradation (both 61 percent). Then there are issues like Interpretability, privacy (54 percent), fairness (51 percent), and safety (46 percent). Interestingly, all these are some form of human issues.

Technology comes with plenty of its own challenges — from deploying at scale to overcoming bottlenecks and skill shortages during development – as outlined by S&P Global Market Intelligence too. 

AI may be moving the needle towards a mint-fresh future – but organizations would need the right talent and the support of relevant capabilities and tools to leverage this potential optimally.

AI will be as pervasive and imperative as air for many enterprises – straddling many industries. But keeping this air breathable and clean would be a job that would need humans – and at a different level of expertise and excellence than ever.

Let’s enter this age of AI – armed with clarity, confidence, and excitement. Having AI by our side would be like having the best breed of dogs – but to make these pets loyal, well-trained and fast – now that would still be a job for humans. Won’t it?