Technology strategy
Intelligent Empowerment: The Next Wave of Technology-led Disruption
Artificial intelligence is everywhere: it promises to power fleets of self-driving vehicles, open up endless new business opportunities and even be the key to global power. Amid all the bluster it’s easy to lose sight of the genuine opportunity AI provides.
For all the current hullabaloo, AI isn’t new. The term “artificial intelligence” was coined back in 1956, at the Dartmouth Conference. It was intended to encapsulate the idea that every aspect of learning, or any other feature of intelligence, can be so precisely described that a machine can be made to simulate it. Fast forward 60-plus years: today, AI has become one of the hottest — and overhyped — tech concepts on the planet.
The burgeoning interest in applications such as speech and optical pattern recognition, picture manipulation, speech synthesis, language translation and self-driving cars has seen a massive increase in research, innovation, and interest. As Google’s AlphaGo showed, with its mastery of ancient strategy board games: AI is now outperforming us in fields that were supposed to be the preserve of human ingenuity.
Little wonder then that tech leaders are exploring what AI can do for them. According to one study, 89% of CIOs are using or plan to use machine learning in their organizations. Analyst firms are typically bold in their predictions too: IDC believes the amount spent on AI will grow from $8 billion in 2016 to $47 billion by 2020.
Companies can use this technology today to gain huge amounts of efficiency, cost savings, and market share. But this isn’t simply about replacing jobs with machines — that will only take you so far. The more compelling opportunities will be realized through the intelligent co-working of humans and machines. We call this Intelligent Empowerment.
Intelligent Empowerment helps people become more efficient, more precise or make better decisions — where intelligent tools augment human capability. Intelligent Empowerment aims to reduce, even eliminate, the boring work humans do today and unleash their creativity and imagination for more fulfilling work.
The availability of massive amounts of computing power on demand allied to advances in machines learning algorithms — and the existence of massive amounts of digital data with which to train these algorithms — opens up the possibilities for Intelligent Empowerment within the enterprise.
The biggest advantage of machine learning systems is the ability to learn from examples. That means you no longer need to define and program the exact behavior of the system — given sufficient training, the machines can work things out for themselves. There have been startling advances in this field, for instance, Google’s AlphaZero, which can learn to play games such as chess by playing itself; within a matter of hours, it’s learned enough to play to a higher standard than the best chess-playing computer.
We already see incredible applications of this technology, in self-driving cars, image and speech recognition, and natural language processing. But we’re also seeing the complexity involved: ensuring there’s enough data available, in suitable formats, to successfully train the systems.
So the tech is best suited to what types of tasks? We think that if several of the following conditions are true, there’s a strong chance that the work could be done successfully by machines:
But as we’ve said previously, Intelligent Empowerment is much more than just replacing human effort with machines: it combines the strengths of both. Machines can tap into immense computational power to process vast quantities of data; humans can use their expertise and ingenuity to do more elaborate work.
Take the global consumer goods company we’ve been working with. To support their planners we built a system to consolidate the company’s operational data and provide a holistic view of the entire supply-chain. Machine learning tech-enabled us to detect problems in real-time and suggest intelligent solutions to address disruptions in the supply-chain, which in turn enabled its planners to increase their productivity 10-fold.
The old system optimized the local supply chains one by one, which worked, in its limited way, but it wasn’t designed to optimize at a global level. To achieve global optimization required a cumbersome re-planning of several local supply chain networks. We designed a system based on the concept of distributed optimization; viewing it as a single global optimization problem that can be solved by mathematically modeling each business objective and constraint with machine learning tech.
We’ve also seen how Intelligent Empowerment can provide innovative new services. For instance, at one international airport, we developed a passenger chatbot to help the 42 million passengers it serves each year. The system helped passengers check their flight status, and could give them directions around the airport, guiding them to shop, restaurants or parking facilities.
Using the principles of continuous development, security and scalability, Thoughtworks was able to deliver the first proof of concept within two weeks.
By dealing with these routine queries, customer-facing staff was free to devote more time to efforts that increased airport revenue. And the chatbot provides a new sales channel for the operator, for instances, offering fliers lounge passes. Unlike its human counterparts, the chatbot can work 24/7, helping the airport to offer continuity of service.
If you invest in Intelligent Empowerment, you’ll want to be mindful of limitations. You’ll also see how fast things are changing, as investments in AI are pouring in.
Currently, identifying use cases for Intelligent Empowerment applications is like shooting a moving target. If your ambitions are too high, you may be setting yourself up for disappointment. Likewise, a lack of ambition and your competitors might create an unassailable lead.
As we talk to our clients, we hear organizations are seeking to understand both the possibilities and the limitations of Intelligent Empowerment, so that they can better plan their roadmap and understand where it can bring them value — and where it won’t.
Next, an enterprise-wide data platform has to be established. This platform must be able to collect, manage and store all defined data streams, which is likely to mean you’ll need to design APIs for existing IT systems and partners. Will you want to create a single data lake or multiple ones? Will the data platform run on your premises or in the cloud? The platform must be able to easily access all data, to store and manipulate that data and to stream it to consuming applications.
What’s more, you’re probably going to have to strengthen the data science skills in your organization. You’ll need a group of data scientists to dig deeply into the data, work with it, understand it and the business background of the data. Easy-to-use tools, like personal notebooks like Jupyter, data analytics, and data visualization tools should be available and in daily use. These data scientists will dig out a lot of valuable and new business insights and decision support out of the data. It will help you to make better decisions based on meaningful data and not just gut feeling anymore.
Finally, the data scientists’ manual work can increasingly be automated by machine learning algorithms. Data evaluation and visualization, and decision preparation can be automated. Machine learning algorithms make it possible to gain even more insights, the behavior of prospects, customers and partners can be learned, your processes can be adapted accordingly, and ultimately, you’ll be able to improve the overall performance of your enterprise.
Simple and repeated decisions all over the company can be automated, the business processes become simpler, faster, more efficient and higher in quality. And by releasing the workforce from simple, repetitive and cumbersome work, a great amount of human capacity will become available for new creative and innovate tasks — what we call Intelligent Empowerment!
For all the current hullabaloo, AI isn’t new. The term “artificial intelligence” was coined back in 1956, at the Dartmouth Conference. It was intended to encapsulate the idea that every aspect of learning, or any other feature of intelligence, can be so precisely described that a machine can be made to simulate it. Fast forward 60-plus years: today, AI has become one of the hottest — and overhyped — tech concepts on the planet.
The burgeoning interest in applications such as speech and optical pattern recognition, picture manipulation, speech synthesis, language translation and self-driving cars has seen a massive increase in research, innovation, and interest. As Google’s AlphaGo showed, with its mastery of ancient strategy board games: AI is now outperforming us in fields that were supposed to be the preserve of human ingenuity.
Little wonder then that tech leaders are exploring what AI can do for them. According to one study, 89% of CIOs are using or plan to use machine learning in their organizations. Analyst firms are typically bold in their predictions too: IDC believes the amount spent on AI will grow from $8 billion in 2016 to $47 billion by 2020.
Companies can use this technology today to gain huge amounts of efficiency, cost savings, and market share. But this isn’t simply about replacing jobs with machines — that will only take you so far. The more compelling opportunities will be realized through the intelligent co-working of humans and machines. We call this Intelligent Empowerment.
More than just intelligence
So is Intelligent Empowerment the same as AI? As we see it, the two are related, but they’re not synonymous. AI is often focused on using computer power to take on tasks that could be done by a person — doing it faster or at a lower cost. It’s can too easily become about replacing humans with machines.Intelligent Empowerment helps people become more efficient, more precise or make better decisions — where intelligent tools augment human capability. Intelligent Empowerment aims to reduce, even eliminate, the boring work humans do today and unleash their creativity and imagination for more fulfilling work.
Figure 1: Machine Learning unleashed
How can Intelligent Empowerment work for you?
“If a typical person can do a mental task with less than one second of thought, we can probably automate it using AI either now or in the near future.”
Renowned AI researcher, Andrew Ng, in Harvard Business Review
Renowned AI researcher, Andrew Ng, in Harvard Business Review
The availability of massive amounts of computing power on demand allied to advances in machines learning algorithms — and the existence of massive amounts of digital data with which to train these algorithms — opens up the possibilities for Intelligent Empowerment within the enterprise.
The biggest advantage of machine learning systems is the ability to learn from examples. That means you no longer need to define and program the exact behavior of the system — given sufficient training, the machines can work things out for themselves. There have been startling advances in this field, for instance, Google’s AlphaZero, which can learn to play games such as chess by playing itself; within a matter of hours, it’s learned enough to play to a higher standard than the best chess-playing computer.
We already see incredible applications of this technology, in self-driving cars, image and speech recognition, and natural language processing. But we’re also seeing the complexity involved: ensuring there’s enough data available, in suitable formats, to successfully train the systems.
Figure 2: Intelligent Empowerment is the application of machine intelligence, to enable new opportunities
So the tech is best suited to what types of tasks? We think that if several of the following conditions are true, there’s a strong chance that the work could be done successfully by machines:
- Clearly defined processes
- It would take just seconds for humans to make decisions
- The processes are data driven, and a large volume of data is available
- Communication with humans or documents from humans are involved — for instance, text, documents, speech or images
But as we’ve said previously, Intelligent Empowerment is much more than just replacing human effort with machines: it combines the strengths of both. Machines can tap into immense computational power to process vast quantities of data; humans can use their expertise and ingenuity to do more elaborate work.
Take the global consumer goods company we’ve been working with. To support their planners we built a system to consolidate the company’s operational data and provide a holistic view of the entire supply-chain. Machine learning tech-enabled us to detect problems in real-time and suggest intelligent solutions to address disruptions in the supply-chain, which in turn enabled its planners to increase their productivity 10-fold.
The old system optimized the local supply chains one by one, which worked, in its limited way, but it wasn’t designed to optimize at a global level. To achieve global optimization required a cumbersome re-planning of several local supply chain networks. We designed a system based on the concept of distributed optimization; viewing it as a single global optimization problem that can be solved by mathematically modeling each business objective and constraint with machine learning tech.
We’ve also seen how Intelligent Empowerment can provide innovative new services. For instance, at one international airport, we developed a passenger chatbot to help the 42 million passengers it serves each year. The system helped passengers check their flight status, and could give them directions around the airport, guiding them to shop, restaurants or parking facilities.
Using the principles of continuous development, security and scalability, Thoughtworks was able to deliver the first proof of concept within two weeks.
By dealing with these routine queries, customer-facing staff was free to devote more time to efforts that increased airport revenue. And the chatbot provides a new sales channel for the operator, for instances, offering fliers lounge passes. Unlike its human counterparts, the chatbot can work 24/7, helping the airport to offer continuity of service.
Staying ahead of the curve
The limitations of today’s AI and machine learning systems have done little to quell the hype. The tech is commonly highly specific; each time you try something new, you’ll have to retrain it. And ultimately, a machine’s ability to make good decisions depends on getting enough quality training data.If you invest in Intelligent Empowerment, you’ll want to be mindful of limitations. You’ll also see how fast things are changing, as investments in AI are pouring in.
Currently, identifying use cases for Intelligent Empowerment applications is like shooting a moving target. If your ambitions are too high, you may be setting yourself up for disappointment. Likewise, a lack of ambition and your competitors might create an unassailable lead.
Figure 3: The holistic approach to Intelligent Empowerment
As we talk to our clients, we hear organizations are seeking to understand both the possibilities and the limitations of Intelligent Empowerment, so that they can better plan their roadmap and understand where it can bring them value — and where it won’t.
Becoming an Intelligent Empowerment driven company
Becoming an Intelligent Empowerment driven company is all about data. As a first step, you have to define a company-wide data strategy and identify important sources and owners of available data. The data owners should be responsible for providing data in real time or at regular intervals, and to a defined quality level. The data owners also need to be aware of the requirements the consumers of their data have. Usually, this needs some organizational changes in the organization. In some circumstances, you may need data from business partners. This requires similar agreements with these partners.Next, an enterprise-wide data platform has to be established. This platform must be able to collect, manage and store all defined data streams, which is likely to mean you’ll need to design APIs for existing IT systems and partners. Will you want to create a single data lake or multiple ones? Will the data platform run on your premises or in the cloud? The platform must be able to easily access all data, to store and manipulate that data and to stream it to consuming applications.
What’s more, you’re probably going to have to strengthen the data science skills in your organization. You’ll need a group of data scientists to dig deeply into the data, work with it, understand it and the business background of the data. Easy-to-use tools, like personal notebooks like Jupyter, data analytics, and data visualization tools should be available and in daily use. These data scientists will dig out a lot of valuable and new business insights and decision support out of the data. It will help you to make better decisions based on meaningful data and not just gut feeling anymore.
Figure 4: We believe that collecting data and building intelligence can only be useful when it’s empowering your business
Finally, the data scientists’ manual work can increasingly be automated by machine learning algorithms. Data evaluation and visualization, and decision preparation can be automated. Machine learning algorithms make it possible to gain even more insights, the behavior of prospects, customers and partners can be learned, your processes can be adapted accordingly, and ultimately, you’ll be able to improve the overall performance of your enterprise.
Simple and repeated decisions all over the company can be automated, the business processes become simpler, faster, more efficient and higher in quality. And by releasing the workforce from simple, repetitive and cumbersome work, a great amount of human capacity will become available for new creative and innovate tasks — what we call Intelligent Empowerment!
Disclaimer: The statements and opinions expressed in this article are those of the author(s) and do not necessarily reflect the positions of Thoughtworks.
