Shailesh Dash


Sunset in the mountains
3 years ago Shailesh Dash

The global deep learning market, currently priced at $4.4 billion in 2020, has been projected to expand at a staggering 39.2 per cent CAGR between 2020 and 2027, to reach $44.3 billion by 2027

Deep learning, a subclass of Machine Learning (ML) and Artificial Intelligence (AI), has been the recipient of profoundly rising interest in recent times. At its core, deep learning technology imitates the way humans gain knowledge, using neural network algorithms to duplicate human thought processes or intelligence for decision-making. The widespread adoption of this technology has been largely abetted by progress in the areas of Big Data, Computing Power and Learning Models. It is estimated that approximately 2.6 quintillion bytes of data are generated every day, which is only testament to the growing importance of deep learning.

Today, the technology has evolved into a powerful tool for organisations, enabling them to gain actionable insights into the expansive range of data they can mine and extract; even unstructured data such as text, speech and images.

Moreover, the exponential growth in computing power, especially through the use of Graphics Processing Units (GPUs), has enabled deep learning to become increasingly feasible, over the past few years. The proliferation of cloud computing and AI as a service, has further accelerated the momentum of this technology, helping it to expand to even smaller companies that can access such algorithms, without making a large initial investment.

Advancements in technology have been accelerating at breakneck pace over the past few years, and most prominently in 2020. Within the tech space, deep learning has been a relatively new, but significant development. The global deep learning market, currently priced at $4.4 billion in 2020, has been projected to expand at a staggering 39.2 per cent CAGR between 2020 and 2027, to reach $44.3 billion by 2027. However, this may only be scratching the surface and potentially witness significant value unlocking during the next decade.

The emergence of the internet era marked a historic milestone with its overwhelming contribution to the global economy, as digital technology created enormous wealth. Its economic potential was even highlighted through the pandemic, as technology-enabled sectors including Information Technology (Google, Microsoft, Apple, etc.), Consumer Discretionary (Amazon, Nike, eBay, etc.) and Communication Services (Alphabet, Facebook, Netflix, Walt Disney, AT&T, etc.) emerged as the most resilient.

Quite similarly, deep learning presents a unique and largely untapped market with immense potential. According to Harvard Business Review, pairing sales with AI technology can increase business leads by 50 per cent, with estimates suggesting that AI could potentially boost average profitability rates by 38 per cent, and incite an economic expansion of $14 trillion by 2035.

Deep learning has contributed to landmark breakthroughs across industries, the most trending applications among which, have been self-driving cars, and chatbots or digital assistants such as Cortana, Alexa, Siri and GoogleNow. While these are just the cream, an expansive range of deep learning applications have been emerging within the fields of medicine, business analytics, hospitality, retail, manufacturing, agriculture, cyber security, automobiles, and more; all of which are experiencing rapid shifts through such applications. For instance, in healthcare, deep learning has been applied to analyze MRI images or x-rays, which have produced results with greater accuracy than human clinicians. In addition, deep learning applications include natural language processing tools, chatbots that can identify patterns in patient symptoms, and algorithms that can identify certain cancers, pathology and even some rare diseases. Aidoc, an Israeli startup that has developed deep-learning based algorithms to support patient diagnosis and treatment in the field of radiology, has received FDA approval for six of its solutions and has raised over $67 million in funding.

Google’s DeepMind is also working to develop a commercialised deep learning tool that can help identify more than 50 different eye diseases with treatment recommendations for each one of them. Meanwhile, in manufacturing, deep learning applications are extended to quality control, predictive maintenance and anomaly detection, which together with other AI technologies such as ML and big data analytics, are aiding a transition to ‘smart’ manufacturing. But perhaps the biggest impact of deep learning is being witnessed in media and entertainment, and the retail industry. Using deep learning algorithms, marketers are gaining valuable insights into customer preferences that are helping them curate influential media strategies and more personalised customer experiences. Rosetta.ai, for instance, uses deep learning algorithms to analyse consumer preferences and help e-commerce businesses increase conversion rates and order value. Alternatively, German startup Luminovo raised $2.5 million to further develop its deep learning capabilities.

The limitless scope of deep learning has prompted tech giants such as Google, Apple, Facebook, Amazon, and Nvidia, to bet big on this technology, and make it an integral part of their product innovation, thereby paving a strong future for this technology.

Alphabet’s open source software library, TensorFlow has become a hub for ML and deep learning software development. Amazon Web Services (AWS) also leverages deep learning, with a focus on marketing that provides more accurate and predictive insights into products that customers might desire. These tech majors have also increased their acquisitions of smaller startups operating within this space, with a growing number of startups reaching advanced funding stages.

SenseTime, one of the world’s most valuable AI companies focusing on computer vision and deep learning technologies, has raised a total of $2.6 billion in funding. Notably, this interest was sparked even prior to the pandemic, with AI acquisitions by tech giants reaching 635 startups between 2010 and 2019. Such acquisitions are enabling the sharing of expertise and massive amounts of data at a global level, the latter of which is a key resource for training models in deep learning.

The enormous potential impact of AI-driven deep learning models on economic development, has also incentivised governments and world leaders to ramp up their adoption of AI technologies. As a result, government entities are gradually turning to deep learning for real-time insights into metrics such as energy infrastructure through analysing satellite imagery, food production, and more.

In 2019, the UK government backed deep learning in healthcare initiatives for the National Health Service (NHS) for £250 million, affirming that AI, ML and deep learning would become a part of government initiatives going forward. US federal agencies also committed $1 billion to AI technologies in the fiscal year 2019. Most notably, China is aggressively pushing for the adoption of AI technologies, to aid its plans to become the world leader in AI by 2030.

Evidently, governments, world leaders, businesses and corporates alike, are recognizing that AI technologies such as deep learning are at the heart of Industry 4.0. While human capacity earlier limited the volume of data that could be processed, the efficiency of deep learning systems is magnified with higher volumes of data, which is likely to enhance data analytics outcomes. Deep learning models are already being deployed across a range of processes including industrial automation, chatbots, text generation, photo/video colourisation, medical research, computer vision and military & aerospace deployment (satellites to detect objects and identify interest areas). Its further potential uses remain unexplored, and remain a key avenue of exploration. Deep learning systems will become easier to deploy on an unprecedented scale, as this technology can bridge the gap between large datasets requiring analysis, and lack of specific algorithms to input. The successful implementation of such systems can potentially disrupt industries and businesses and change the world as we know it. The Covid-19 pandemic, which brought economies to a standstill, only accelerated the process of global digitization. Consequently, deep learning has assumed greater importance, as governments and corporates unanimously push for the adoption of such technologies that can empower their workforce and give them a significant competitive edge.