Quanton's World of Business Process Automation

San Francisco AI Summit: Key Learning and Takeaways

31-Oct-2019 08:50:52 / by Russell Berg

Ai Summit San Francisco Key Learning and Takeaway

 

 

Automation is undoubtedly on the agenda of most executives in New Zealand. Enabled by a wider ecosystem of smart technologies, Intelligent Process Automation is the destination making AI one of the hottest words of 2019.

In September I travelled to America to attend the San Francisco Ai Summit – self-proclaimed the world’s largest AI event for business. At the event I joined several thousand attendees to hear from over 200 speakers and engage with over 200 exhibitors.

The aim was to bring back key learning to be translated into practical insights to share and help move New Zealand businesses forward in the adoption and application of smart technologies – something key to Quanton’s drive.

 

 

 

We strongly believe our purpose is to help New Zealand enterprise create new ways of working. Quanton does this by driving transformational change within the business to create future focused operating models that balance people, processes and technology.

 

 

 

 

There is an array of technology under the banner of ‘AI’ and differentiating between the hype and practical reality can be difficult. As a result of the ‘newness’ and rapid advancement of AI technologies, effort can be disproportionate to benefit, creating large barriers. Several technologies such as Machine Learning and conversational AI are now generally viewed as being past the hype and well positioned to deliver business value across the total enterprise.

Global leaders have a scale, which is incomparable to New Zealand. That scale means it can be easier to create business-cases for, implement and scale advanced capabilities – which is one of the aspects I was most excited about.

But despite that scale, I came away from the summit with the strong feeling that New Zealand businesses are keeping up with the best in the world in the way we’re developing capability and in where our organisations are at in their journey.

 

 

Advanced enterprises are preparing to scale capability across their operating models.

 

 

One piece of research, AI Transforming the Enterprise – Eight Key Adoption Trends, presented at AI Summit, highlights the struggles companies are experiencing to scale their technology capability and programmes across the total enterprise. This is a common challenge which I believe we also see in New Zealand.

 

 

San Francisco Ai Summit Key 8 AI Trends

 

 

 

For example, the research report, presented by KPMG, shows 26% of respondents have deployed Robotic Process Automation, however over half of these respondents (65%) say their use is selective and siloed; And only 17% of respondents interviewed reported the use of AI or machine learning at scale, while 30% reported use in selective functions.

But 83% of respondents expect to have RPA deployed and half of the companies interviewed expect to be using AI and machine learning at scale. This shows a clear desire to scale AI programmes – but that’s easier said than done. One of Quanton’s key messages, which these figures support, is that true transformation can only occur when people, technology and processes are considered in balance. There is a critical dependency on the relationship between the business and technology, and more so on the engagement of the total population in the organisation.

A quote which I heard at AI-Day (Auckland 2017) by Steve Guggenheimer, Corporate Vice President AI and ISV Engagement - Microsoft has never been truer:

 

 

"It's too early to do everything, but it's too late to do nothing." 

 

 

In the following article I share a summary of the top 10 takeaways from the conference. Many of these are significant topics which I will explore fully in follow-up articles and podcasts over the next few months.

 

 

Key Takeaway #1: Addressing ethical issues is central to the future enablement of AI in businesses

 

Ethical issues including bias, privacy, access and transparency are top of mind. Right now, it seems like we are 85%-90% reliant on self-regulation.

If we want to understand the potential risk, we only have to look at the example of Cambridge Analytica, who are accused of using data to influence political outcomes in the United Kingdom and America.

But despite its importance, it’s a conversation that hasn’t evolved much at all in the past two years. We’re still hearing the same discussions, with little action or movement forward. And that’s a big issue.

There is a consensus that we need more robust frameworks and legislation to protect people, both locally and globally, but the application, utilisation and adoption of advanced technologies are moving faster than the conversation for ethics can take place.

Until we get to the point where we have clear legislation and frameworks the only thing, we really have is transparency. I think transparency is one of the greatest areas where legislation has a role to play in relation to AI and smart technologies.

This is a massive discussion that I start to explore in the first follow-up article, The Ethics of AI, which has been published by CIO.co.nz.

 

 

Key Takeaway #2: If customer experience is a strategic advantage, the realm of technology enabled possibility is limitless

 

We’ve been talking customer experience (CX) for a long time, but in the past customer experience has been constrained by what is possible through the current boundaries of people, processes and technology.

When you consider the advancements that have been made such as machine learning and automation, those constraints are now being blown out of the water.

But here’s the thing to consider: You are not going to be measured by how you rank alongside your immediate competitors – instead, you’re going to be measured by best in class experience, whether that best in class is in your industry or one completely separate, be it finance, energy or something else.

The historic approach to customer experience design and management has often been to design something around customer experience and then… well, leave it be. But placing the customer at the centre of the journey means continuous iteration and review. Customer expectations and needs are constantly evolving so your customer experience has too.

Through technology we can collect and use data and gain a real-time view of the customer.

Don’t look at the customer in a vacuum, look at data and understand how it can add value in context, in the present. With the capability we now have around machine learning and the ability of algorithms to self-evolve and self-learn, if your customer service is built around algorithms there’s a very strong opportunity for your service to also be continually evolving. In the future, it’s quite likely we’re going to see fully self-evolving services.

It is likely that businesses who ignore the CX will be left behind. The key message for me: As we move into an era of personalisation, predictive analytics and automated experiences we have to be clear on who we are serving and where they are in the customer journey, to provide the customers with any combination of offers, information, service or utilisation. All in a manner that is highly relevant, contextual and adds value at a given point in time.

 

 

Key Takeaway #3: People are more important than ever

 

The culture of an organisation must support a technology-enabled approach. We need to get away from thinking of these projects as purely technology projects. Instead understand they are change projects – you are fundamentally changing how your business operates. And that change will fail unless everyone is on board, engaged and supporting the moves.

People must empower the technology. There is no compromise.

It must start with the leadership and run right through to the front line, placing more emphasis on change than ever before.

This can, however, be a challenge for organisations, particularly when the organisational structure has multiple layers. We often see the top level of management on board, but disparate messages being heard in lower levels – yet these are often the most important people to have on board as they are the ones interacting with customers.

Perhaps most aptly described in a presentation I saw by Anthony Scott, whatever you estimate for change management, go outside, take a breath come back and double it.

 

 

Key Takeaway #4: Data is fuel

 

According to The Great Hack, Netflix’s documentary on Cambridge Analytica, last year data surpassed oil in value. Data is becoming one of the most valuable assets a business will have.

The success of an organisation, relating to the deployment and application of AI technologies, will be based on their ability to capture, call, store, process and access data. Equally, AI creates ten-fold more data.

We all know the old saying, which for politeness sake I’ll paraphrase as: “Rubbish in, rubbish out”.

Data is a critical dependency. The commercial and ethical risks businesses expose themselves to with poor data, especially in relation to algorithms are real, not only in terms of reputation impact, but commercial outcomes.

You need to understand what data you have in your business and the value of what you have – something I don’t believe many companies are leveraging yet.

Think about how you can make your data available in a centralised way – how can you draw from data via a single centralised place, and see it and visualise it in a one place, how can you move it, store it, integrate it. Thinking about your framework around data is a massive piece of the AI puzzle because it is going to become a unique piece of infrastructure.

 

 

San Francisco Summit Key Learing Data Hierarchy

 

 

 

Key Takeaway #5: It's time to adopt organisational structures that support our future operating models

 

Current organisational design is based on old ways of working and a higher dependency for human labour, but the more we harness technology, the less relevant that becomes. Organisations are going to have to reconsider the structures that will be support their future-ready operating models and new ways of working

As technologies such as machine learning, conversational AI and intelligent automation begin to be applied at scale we will see the continuous iteration of services and digital products. As machine and deep learning, continue to advance, we will also see self-improving models, services and products.

At the same time, all the businesses I speak to know the environment around them is changing and they need to change faster – but they’re constrained by their own internal processes.

The traditional structures and processes of large enterprise, which are largely based on control, will become prohibitive to an organisation’s ability to keep pace with environmental change.

Organisations which want to evolve will have to assess how they are structured. The suggestion is that organisations will have to find ways to decentralise control and decision making, placing a greater level of empowerment in their teams.

The question is how are you going to empower your people more, decentralising decision making so it’s closer to those at the coal face?

 

 

Key Takeaway #6: New approaches are required to avoid a wave of unemployed

 

Automation and robots, both physical and digital, are changing the world and we are at risk that automation will create a negative social divide.

How Robots Are Changing the World, a report by Oxford Economics which was presented at the conference, claimed that the number of (physical) robots in use worldwide multiplied three-fold over the past two decades, to 2.25 million. Trends suggest the rate of robot use will increase and as many as 20 million robots could be in use by 2030.

Three years ago, the media was awash with stories about how we were all going to lose our jobs to robots. Today however, it’s generally accepted – and backed by research from the likes of The World Economic Forum, Price Waterhouse Cooper, McKinsey, The London School of Economics and the recent report by Oxford Economics – that while we will see job loses, we’ll also see the creation of new jobs, with a general consensus that those figures will even out.

The jobs displaced, however, are low-skilled jobs compared to the jobs which are created, which are generally assumed to be medium and high skilled jobs.

This raises two inherent problems. The first, which I will only touch on, is that of a skills shortage and the second is that we are staring at the potential for a new wave of low-skilled unemployed people. That’s something that could increase the divide between those who have, and those that have not.

The three parties directly involved in all this are the technology providers, employers and the employees. 

But other parties – educational providers and governments – also have a role to play.

There are some tough questions to explore around what responsibility will look like for each of those parties and what level of focus we place on initiatives over time to mitigate the impacts of this potential problem.

Whatever happens, the time to start looking at this issue is now. How do we up-skill those at risk of losing their jobs now? We can be the fence at the top of the cliff – or we can wait to supply all the ambulances at the bottom of the cliff later on.

 

 

Key Takeaway #7: Adopt an open approach to architecture, to create best-in-breed capability

 

AI is a category of technologies. Best in class digital transformation and the operating models of the future will not be achieved through one technology. It will be the application of multiple technologies in aggregation which will create the future opportunity for businesses.

This means companies need the ability to bring together the mix of technologies that best suit their business, service delivery and operating model.

Keeping an open architecture with a modular approach to technology, enables organisations to develop unique capability for their business and provides ultimate flexibility, enabling them to swap out anything no longer serving their business as the business needs change. Furthermore, there is a clear move away from developing solutions towards developing platform-based capability. This is dependent for scalability.

 

 

Key Takeaway #8: The ability to integrate systems is a critical dependency

 

Businesses have to address the challenge of integrating a wide variety of technology from legacy systems to specialist applications and web-based services.

Enterprises have huge volumes of technologies in their business, providing challenges from a support, cost management and licensing perspective, among other things, and simplification has been the name of the game for most IT teams for a while.

AI however, goes against that. In most cases AI technologies are not replacing existing technologies. They are additional technologies within the wider technology mix exasperating an existing challenge.

One of the key dependencies for businesses to be successful in the era of AI will be the ability to integrate systems and applications to support the flow of information and execution of processes.

I would go as far as to place this second only to data as one of the big pieces to resolve. The takeaway - Businesses have to now develop a suite of capability around integration. Organisations will need several highly flexible integration options – a space which Robotic Automation fits into.

 

 

Key Takeaway #9: Build organisational capability, not solutions 

 

There are a large number of organisations who have initiate ‘Ai’ programmes however these are in the often in the form of point in time solutions.

The opportunity is for businesses to build capability based on people, processes and technology which enables the constant evolution of Ai to better meet business needs and improve the benefit which can be derived.

Focusing on technology, a clear message from many of the speakers was the shift away from solution orientation to platform capability. One of the most prevalent areas which came through at the Ai Summit related to conversational AI and the development of conversational platforms that were agnostic of channel. For example, a conversational platform which could then be integrated to any number of relevant channels like Facebook Messenger, SMS and email, instead of building point solutions like a chat bot.

 

 

People, not technology are critical to the success of capability.

 

 

KPMG claim the largest enterprises in America are making major investments in people and plan to extend that investment. The top five firms with mature AI capabilities interviewed for AI Transforming the Enterprise – Eight Key Adoption Trends, had on average 375 full time employees (FTE), which includes but is not limited to data scientists, engineers and analysts.

KPMG estimated that on average each of these enterprises were spending $75M on AI talent, and they expect to continue growing to between 500 and 600 FTE working on AI in the next three years.

Capability starts with clearly defined strategies at an executive level and is enabled by the right structures, leadership and oversight – the ability to find the optimal balance of people, processes and technology.

It is capability that will enable businesses apply smart technologies at scale and create the operating models of the future based on new ways of working.

 

 

Key Takeaway #10: Strategic governance is a dependency to achieve enterprise-wide application of Ai capability

 

The necessity for governance is inferred throughout the previous takeaways especially around organisational capability, scalability, organisational structures and ethics, but a dedicated takeaway is still warranted.

Governance is not a new concept to organisations but existing governance structures and processes have not necessarily been adapted to support enterprise to embed and scale capability across the business.

Governance around smart technologies include aligning benefit outcomes to strategic drivers, benefit realisation, designing and deploying standard procedures around monitoring and managing risks, performance and value, ensuring end-to-end processes maintain trust and transparency and determining roles, responsibilities and accountability. Perhaps most importantly managing change, a topic covered in a recent article, when an automation programme fails, this is usually why – by Adam Taylor.

 

 

Summary Left

 

In summary, each of these ten takeaways, are significant conversations, many of which I intend to explore in future articles.

By way of closing I would leave you with one final thought taken from a presentation on automation predictions for 2020:

 

"You need to figure out how to inculcate automation and AI in your business from today, in order to be able to adapt to tomorrow."

Guy Kirkwood, Chief Evangelist, UiPath

 

Summary Right 

 

 

Topics: Digital Transformation, AI, Artificial Intelligence, Intelligent Automation, Future of Work, Cognitive Automation, Machine Learning

Russell Berg

Written by Russell Berg

Drawing on experiences from the public, private and not-for-profit sectors Russell draws on a diverse background, exploring how smart technologies are enabling capability for enterprise businesses.