What is the AI in your strategy? Artificial Intelligence or Almost Impossible?

“Culture eats strategy for breakfast” – famous words once uttered by Peter Drucker and made popular in 2006 by the ex-president of Ford, Mark Fields. A brilliant strategy is doomed to fail if the culture of the organisation is unable to support it.

My career in data and analytics, delivering change across multiple organisations, has highlighted the importance of engaging staff to embrace data-driven change. Fail to do so and it undermines the entire programme.

So, what is the connection with artificial intelligence (AI)?

Academia recognised AI at a 1956 Dartmouth College conference, though the concept dates back much further. Banks have used AI since the 1980s, yet today’s AI hype makes it seem relatively new.

The pace, range and impact of applications is increasing visibly. From the cars that we drive to the computer games we play, recruitment to medicine, AI is increasingly influencing our decision-making.

Major global players are investing heavily in AI. In 2015, Sundar Pichai, Chief Executive of Google, announced Google was to become ‘AI first.’ Whilst the implications were complex, fundamentally it recognised the future lay not with traditional computer programming but with machine learning.

Google Brain, founded in 2011, has tackled the most challenging applications of technology – image recognition, language translation, speech recognition – at a phenomenal rate. Today, neural networks achieve results in some of the most complex and specialist human tasks. For instance, a radiologist requires years of expensive, specialist training to detect tumours through medical images. AI detects tumours earlier but can also identify via pathology reports. AI cannot find the cause of the tumour, but it can replicate what the radiologist is doing.

AI is so prevalent that we live with it, subconsciously, daily. Yet there is growing discomfort at how commonplace it is becoming and ethics surrounding its use.
In Shenzhen, China, a facial recognition system was implemented in 2017 to deter jaywalking. Initially, an image, the person’s name and part of their government ID number were displayed on a large screen by the intersection. It was expanded, with individuals receiving a mobile phone text message notifying of a fine almost immediately. This is part of the social credit system, rolling out in 2020, with individuals scored based on how trustworthy a citizen each person is considered.

There are concerns those with a low score might be refused certain jobs or pay more for certain services.

Google’s code of conduct included the phrase ‘Don’t be evil,’ but this was replaced with a more opaque reference to ‘ethical business conduct.’ Whilst ‘don’t be evil’ remains at the end of the code, the Google-US military relationship has Google staff concerned about AI being used to develop autonomous weapons. Google announced it will not undertake such activity but will progress other AI initiatives with US military. The US military intends to exploit AI and others will fill the void.

Clearly, AI transcends our daily lives, but the ethics of AI raise challenges and are likely to be felt by businesses through their customers and staff. Many question who regulates AI, especially given the pace of development.

Delivering organisational change is challenging. People are often resistant to change due to the unknown or, potentially, a known scenario where risks are perceived to outweigh benefits. AI-driven change is hard to sell as it lacks emotional intelligence to articulate reason and direction and is ‘black box.’ Deviation from the direction specified by AI increases risk of failure.

49% of AI strategy implementations fail due to cultural resistance according to a new research report: "AI2020: The Global State of Intelligent Enterprise."  Subject matter experts (SMEs) must be engaged to succeed, yet they may feel most threatened by it. Advocates want to understand the drivers of change, and an inability to articulate these increases discomfort. Communication is essential, but without the emotional intelligence and compelling story it is difficult to convey a compelling message.

In late 2017, Gartner polled CIOs about AI in their businesses . Only 4% claimed to be using AI and 85% of AI projects were deemed failures. Further investigation illustrates the importance in getting the basics right.

Firstly, most organisations do not understand the importance of data - its availability, accessibility and, fundamentally, its quality. Sherlock Holmes exclaimed ‘Data! Data! Data! I can't make bricks without clay.’  Data is often missing, erroneous or lacking updates, all of which are critical in building an intelligent organisation. This undermines AI from the outset.

Secondly, most lack the AI talent to exploit it. Leading US digital companies go to extraordinary lengths to secure top AI talent, with their owners making personal calls to individuals. Most traditional bricks and mortar organisations lack understanding of AI’s potential, let alone spotting and attracting top talent, harnessing and empowering it to change corporate strategy. Some suggest innovation hubs or subsidiaries with a blank canvas or test bed are more likely to succeed than transformation from within.

Google Brain had the benefit of five years of development in such an environment, and having honed their AI capabilities, saw an opportunity within the Google Translate product. They ran the existing programmatic method alongside the new AI approach, translating English to French and saw improvements roughly equivalent to the aggregate gains of the old system over its entire lifetime in just a few short months. They quickly stopped development of the old system and focussed entirely on making the new AI approach robust, scalable and able to cover 100 languages.

Thirdly, some say AI’s the latest CIO technological fashion accessory. Andy Rowsell-Jones, vice-president at Gartner, suggests caution as it may simply be an ‘irrational exuberance’, responding to the industry’s own push towards what AI may offer, in other words a belief that they should adopt AI because everyone else is.

Subsequent Gartner research waves with CIOs on AI may substantiate this. Earlier this year, the same Gartner annual survey  saw growth of 37% year-on-year and Chris Howard, research vice-president at Gartner, stated ‘we have now entered the realm of AI-augmented work and decision science — what we call ‘augmented intelligence’.' However, widespread adoption and AI driving significant change at a programmatic level still seems some way off.

AI is a growing phenomenon with familiar problems – data, people and process tripping up smart technology. Regulation will probably determine boundaries eventually, and AI faces many familiar cultural battles. AI will not flounder - it is already too established - but the journey ahead looks both turbulent and exciting in equal measure!

Ian Wallis
Data Strategists Ltd
June 2019

Ian will be the keynote speaker at our upcoming event "Exploring the ethics of AI in today's workplace." on Thursday 26 September, 2019. If you're interested in attending, please email us to register your interest.

References

1.  The Great AI Awakening, The New York Times, 14 December 2016
2.  IQPC's Intelligent Automation Network.
3.  2018 Gartner CIO Agenda Survey
4.  Sherlock Holmes in ‘The Adventure of the Copper Beeches,’, Arthur Conan Doyle (1892)
5.  2019 Gartner CIO Agenda Survey


We use cookies to give you the best browsing experience. This may include storing your login information and using third party cookies to understand how you use our site. By clicking Accept, you agree to the use of cookies. See our Cookie Policy for more information.

Accept