Highlights from McKinsey’s report: “AI - The Next Digital Frontier”
McKinsey is becoming a prolific signposter of automation trends, so we thought it both timely and important to highlight the key elements of their recent report: Artificial Intelligence, The Next Digital Frontier.
The report analyses the AI investment landscape, the portfolios of internet companies and a survey of over 3,000 C-level executives’ views on automation to assess the potential impact of artificial intelligence.
The increasing importance of automation
The first key theme of the report is that automation driven by AI is one of the most transformative processes affecting the global economy today.
McKinsey state that in 2016, twice the number of articles mentioning AI were published compared to 2015, and four times as many as in 2014, signalling a marked increase in interest.
According to McKinsey’s estimates, the combined AI investment landscape rose from $26 billion to $39 billion in 2016.
McKinsey have identified distinct themes that have been pivotal to the recent surge in AI investment:
- An overwhelming majority of current investment in AI research is coming from technology companies themselves, with strong R&D teams ready to test and deploy AI solutions
- A rise in the velocity, variety and volume of big data available provides a rich source of data for machine learning systems to parse, model and subsequently utilise to increase automation and efficiency
- The above has been compounded by an increase in computing power, mainly centred around the development of faster, more advanced graphics processing units (GPUs)
More firms are beginning to invest in AI, but many are missing out on its potential
Of 3,073 C-level respondents across the health care, retail and telecoms sectors:
- 20% of businesses have adopted one or more AI solutions either at scale or have incorporated it as a core element of their operations
- 10% reported introducing more than two artificial intelligence products into their core business model
- 9% reported using machine learning to model their data
These numbers are still exceptionally low when pitted against the potential returns. Even at this early stage of human-to-computer workflow integration, failure to invest in AI will be costly longer term, as the huge gains made by early adopters illustrate.
- Amazon improved their click-to-ship cycle time from a human performance of between 60–75 minutes to less than 15 minutes using robotics. This had a compound result of reducing operations costs by 20%.
- Netflix improved their notorious recommendation algorithm, reducing their yearly churn rate by $1bn.
Similarly, Echobox has increased Facebook traffic for publishers by 57% on average, and has saved publishers over $15 million per year, by integrating artificial intelligence into the workflows of their social media teams.
Early adoption is key
20% of businesses from sectors as diverse as technology, healthcare, retail and telecoms have adopted one or more AI-related technologies as a significant element of their core business.
Early adopters of artificial intelligence largely shared several key characteristics. They tend to be digitally mature, in that they are already investing at scale in new technologies such as big data. Early adopters also tend to be larger businesses that have adopted AI at scale and use it as a core part of their business. In addition, McKinsey observed that these corporations are more motivated to invest by the promise of service innovation than financial gain alone. Finally, C-level support for AI was found to be a key factor determining whether there was a successful and early integration of artificial intelligence.