Machine learning to intensify: Deloitte

Advancements in machine learning technology include data science automation and a reduced need for training data

March 19, 2018
Purchasing B2B

TORONTO—Machine learning will intensify amongst medium and large-sized enterprises, doubling the number of implementations and pilot projects using machine learning technology in 2018 compared to last year, and then doubling again by 2020. According to Deloitte’s Technology, Media and Telecommunications (TMT) Predictions, advancements in machine learning technology include data science automation and a reduced need for training data as well as new chips in both data centres and mobile devices. The advancements will help establish the foundation, which will over the near term make machine learning mainstream across industries where organizations have limited talent, infrastructure and data to train models.

Machine learning is an artificial intelligence (AI), or cognitive technology that enables systems to learn and improve from experience—by exposure to data—without being programmed explicitly. Deloitte predicts that by the end of 2018, more than 25 percent of all chips used to accelerate machine learning in a data centre will be field programmable gate arrays (FPGAs) and application-specific integrated circuits (ASICs). These chips, in addition to graphics processing units chips (GPUs) that were the main choice in the past will dramatically increase the use of machine learning, enabling applications to consume less power while being more responsive, flexible and capable.

Now in its 17th year, Deloitte’s annual TMT Predictions provides an outlook on key trends over the course of the next five years in the technology, media and telecommunications industry sectors worldwide. Deloitte’s TMT Predictions are based on global research including in-depth interviews with clients, industry analysts, global industry leaders and Deloitte member-firm TMT practitioners. Last year, Deloitte was more than 85 percent accurate with its TMT predictions.

The most significant TMT predictions to impact the Canadian marketplace in 2018 include:

Machine learning: In 2018, the number of machine learning technology implementations and pilot projects by medium and large-sized enterprises will double compared with 2017, and double again by 2020.

Hitting the accelerator: By the end of 2018, over 25 percent of all chips used to accelerate machine learning in the data centre will be FPGAs and ASICs. The new semiconductor chips, in addition to GPU chips that were the main choice in the past, will increase the use of machine learning and enable applications to use less power, yet still be more responsive, flexible and capable.

Live thrives in an online world: Live broadcast and events will generate over US$545 billion in direct revenues in 2018 globally and C$17.5 billion in Canada. Despite consumers’ capability to consume content on demand or attend events remotely, live consumption is thriving and can be made more productive and profitable by digital.

Smartphones are useful, but they can be distracting: By the end of 2023, more than 90 percent of Canadian adults will have a smartphone, with ownership among 55-75 year-olds reaching 85 percent. Owners will interact with their phones on average 65 times per day in 2023, a 20 percent increase on 2018. However, 39 percent of Canadian adult smartphone users and 65 percent of 18-24 year olds will worry that they use their phones too much for certain activities and will try to limit their usage in 2018.

Digital media: well worth paying for: By 2020, there will be over 680 million digital subscriptions, with consumers increasingly willing to pay for content.

Augmented reality on the cusp of reality: Over a billion smartphone users will create augmented reality (AR) content at least once in 2018, with at least 300 million doing so monthly, and tens of millions weekly. This translates to about 15 million Canadians using AR.

The future of the smartphone: Smartphone sales will reach 1.85 billion units per year by 2023, equivalent to over five million units sold per day. The smartphone’s success over the next five years is likely to be the introduction of innovations that are largely invisible to its users.

In flight connectivity takes off: One billion passenger journeys, or one quarter of all passengers, will be on planes fitted with in-flight connectivity (IFC) in 2018, an estimated 20 per cent increase from projected 2017 totals, generating IFC revenue close to US$1 billion for 2018.