As a preparation for the AI 4 Business Summit, we’ve interviewed our AI experts.
Here you can find their insights as well as other related content.
Be sure to visit this page regularly as it will be often updated with new posts.
It’s clear: Artificial Intelligence is the next big thing. Not just a technology hype, but here to stay. We’ve asked our main partners to answer the three most commonly asked questions about AI. First up: Microsoft.
As one of the worldwide leaders in software, services, devices and solutions for business and people, Microsoft was one of the first pioneers to start experimenting, researching and implementing Artificial Intelligence to accelerate businesses. They are a firm believer that AI has the capability to amplify human ingenuity.
Mrs Myriam Broeders, CTO of Microsoft Belux kindly answered our three questions. She’s also one of our Keynote speakers at our summit.
KOOACH // How do you see AI evolve in the next three years?
Myriam Broeders // “The current democratization of AI – making it broadly available for companies and consumers – will lead to a giant leap in AI infused applications. This will accelerate the learning of systems and bring more and more accurate outcomes, meeting or even exceeding some human skills, like the trend is already there for e.g. image and speech recognition.
AI and machine learning will continue to impact our business processes and interactions, within and between companies, with customers, and as consumers. It will enable us, at the right time, with the right level of information, to act faster, to take better decisions faster and to manage our time differently.
At the same time, we will have to consider societal concerns related to AI: bias, machine mistakes, unintended consequences, … As stated by Satya Nadella: “The most critical next step in our pursuit of AI is to agree on an ethical and empathic framework for its design.”
We need an ethical and empathic framework for AI
KOOACH // What’s the biggest advantage for AI for business?
Myriam Broeders // “’AI for business’ unlocks potential that was not reachable before. It allows us to go where no man went before, amplifying human ingenuity with intelligent technology. Think of drones that via cameras and image recognition are able to verify high-voltage cables. Think of the time given back to people in offices, having a digital assistant managing agendas. Think of better financial forecasting and trending through AI and machine learning. Businesses are able to respond in real-time, anytime and anywhere to customer’s needs and can do more than before was thinkable or possible.”
AI unlocks potential that was not reachable before
KOOACH //Which industries benefit the most from AI currently?
Myriam Broeders // “Technology, (financial) services and retail companies have already adopted AI in a number of their base and core processes. In the industrial area, more and more the benefit of AI and machine learning is recognized, and applications & implementations start delivering great examples of cost and risk reduction, of predictive maintenance and production optimization. In healthcare and life science, R&D is investing heavily in AI and machine learning in order to improve diagnose, medicine usage, patient follow up but very importantly also in disease analysis and curing options.
Governments are already engaged in AI projects for smart cities, energy savings and resource management. They also look into more sensitive services, like digital ID with face recognition and like image recognition via street cameras. Ethical considerations, human monitoring and constant feedback loops, will become key in taking the right decisions for AI implementation.”
Curious of what other insights Mrs Broeders and Microsoft have to share? Follow her Keynote about why you don’t need Artificial Intelligence to start with AI at 11h30 in Studio 16. Haven’t bought your tickets yet? Grab your entrance passes here.
We constantly interact with our environment by using all our senses. A changing environment will lead to the creation of new concepts in our head which will be expressed through new words. AI is no different in this as was the Cloud, and the Internet before it. AI is becoming a very strong change factor so AI related terms will be incorporated in our vocabulary and become the new standard.
The three most heavily used AI- topics by nearly anyone (including business context), without people being aware Artificial Intelligence is involved are: email categorization (including SPAM filters), search engines and recommendation engines. These are just some examples of how AI is capable of easing our day-to-day lives.
The following implementations are not omnipresent yet, but are on the rise: machine monitoring and control systems, digital manufacturing (by helping reduce cold-start loss, quality control,…) and intelligent ERP-processes (for example fraud-detection, payment matching, CV matching, smart procurement,…)
Do you need to know how to code to build landing pages? 5 years ago: yes. Today, not as much. Tools like Instapage, Unbounce or Strikingly allow us to build landing pages without needing any vanilla HTML.
The typical business applications of machine learning, such as predictive modelling and clustering, are relying less than ever on the production of original code. For instance, in AI software-as-a-service like Dataiku, MS Azure and Orange, the code is hidden under the hood of easy to use drag and click interfaces.
The first reason is the rise of MLaaS (Machine Learning-as-a-Service). The rationale for it comes from the fact that: 20 percent of the tools in a toolbox are used in 80 percent of tasks (Pareto’s Principle). The remaining 20 percent of the tasks probably require customised code.
Both judging and creating art are well established applications of Artificial Intelligence. Machine learning models are so accurate that they can judge the authenticity of paintings by merely looking at its brushstroke patterns. Algorithms for predicting the popularity of songs have been around for more than 10 years. Music generation software has been around for decades and it is now trivial to generate symphonies in the style of Bach, Mozart or Metallica. Recently, Deep Dream has shown us how to apply art styles to photos and generate funky new compositions.
Pearl, the first AI jury member in the world, recognises this phenomenal accuracy of machine learning models in judging creativity. She takes the models to the next level by using them to replace and augment the human judges. With success: the human judges ended up picking the exact same winner as Pearl. The main difference is that Pearl came to her conclusion in a matter of seconds, where human judges had lengthy meetings and took multiple days to come to the same – perfectly predictable – result.
AI is getting closer and closer to human-level performance. In some areas human performance is already beaten.
In areas like in a Call Center or a Shared Service Center, where we see repeating tasks, AI can increase productivity and improve customer experience because response times will get shorter and accuracy will get higher. At SAP we have started embedding Machine Learning models into our Enterprise Software to make our applications more intelligent. This results in higher automation and faster end-to-end processing of complex ERP processes. A good example you can see here in the area of processing invoices and payments.
Departments leading in the use of AI and machine learning, impacting revenues and cost, are finance for fraud detection and forecasting, sales for upsell propensity and churn reduction, and marketing for best offers and return on investment insights. Also HR and IT departments introduced AI and machine learning in their services, with intelligent bots addressing employee questions and problem resolution. In digital communication with customers, via smartphone apps and on websites, many companies introduced bots as great conversationalists, supporting commercial activities from marketing to Q&A and sales. Introducing AI in business & industrial processes can and will lead to disruptive new ways of doing business, of changed employee and user experience, and of new ways of creation of things and services.
It is very interesting to see how the definition of Artifical Intelligence is heavily influenced by the progress that technology makes, and especially the direction in which the research progresses fastest or easiest. We are shaping the meaning together with the concept itself.
At this moment, for AI systems being implemented nowadays, I would mainly describe Artificial Intelligence as increasingly complex pattern recognition, and the use of this ability to predict future behaviour or events. Maybe you could say that AI starts when systems move away from the linear progressive calculation that computers mainly have been doing, and start to branch out, and evaluate the different branches.
For the future I think Artificial Intelligence will be about creativity. Then we have to first found out how our own creative thinking works. My guess is that creativity as a brain process has two essential aspects: the generating of many different ideas/concepts/possibilities in as many directions as possible, and having a good evaluation system that both knows what a “good” result would be, and that can constantly evaluate the different directions in which the system is thinking. Evaluate the developping branches too soon or too harsh, and you’re back to linear calculations. Evaluate them too loosely and you’ll spend your computing power generating a large set of vague ideas and concepts without direct use.
As markets evolve, efficiency alone will not keep businesses thriving. By providing context-relevant information, predictive insights and constant learning at every point of work, businesses can increase proficiency in real time, throughout the enterprise and beyond. Proficiency will become the essential competitive attribute, the hallmark of agile businesses.
Powered by AI, the intelligent workplace will filter incoming information based on our preferences, inform you on approaching deadlines and observe and report on our messaging behavior.
Using cognitive functions, intelligent workplace will lower language and speech barriers between workforce members. It will classify produced content (instead of the creator) and aid and speed up information discovery.
The intelligent workplace will convert voice commands into actions across multiple platforms, learn continuously and will create new associations between people, activities, documents and business data.
Finally ,the intelligent workplace will present smart dashboards that show real-time information, predicts issues, prevents problems and initiates counter-actions.
According to Sven Arnauts, who is responsible for Performance Management & Customer Insights at delaware BeLux, allowing AI to get into the core of an organization’s corporate performance management cycle enables to take full advantage of all data available in the organization’s ecosystem in order to take the best possible decisions in previously unseen ways.
My top three AI-Startups are:
Performance Management has always been a reactive reporting process, and to some extent, was trying to predict the future. When we look into KPI-measurement, or into profitability analysis, or strategy management,…we are always driving our decisions based on historical information and trying to sense which direction to go to. Artificial Intelligence is an enormous opportunity to create Performance Management 2.0. Allowing artificial intelligence into the core of your corporate performance management (CPM) cycle will enable you to take full advantage of all the data available in your ecosystem. Detecting patterns, performing predictions, analyzing large amounts of market intelligence,…all these activities have the objective to look into the future and show you the things you’ve never noticed. So yes, AI will have a huge impact on performance management because it will allow us to correlate information and create insights that weren’t possible before, so we can take the best possible decision.
At this time, many countries do not have legislation that specifically targets AI systems and their implementation in day-to-day life. The UK, however, has a parliamentary committee “to consider the economic, ethical and social implications of advances in artificial intelligence, and to make recommendations.”
Some situations could be dealt with with current laws and regulations, but it could become an uncomfortable exercise. That is why it’s of the utmost importance to get AI on the lawmakers’ agendas sooner rather than later.
Moreover, the proactive stance of the lawmakers would be in contrast with the way technology and law have evolved in the past decennia. On many occasions, the developments in the technologic world have imposed their ethical standards simultaneously or at a later stage, when abuse became apparent. Lawmakers have in general been late to the party, which is why I think it would be a good thing to implement such a government official in every country to analyse the impact of new technologies on different aspects of society, and set the balance straight before abuse happens.
The focus of the sports analytics community has been very much on collecting more and more data during matches and training sessions. However, what is even more important than collecting the data, is extracting actionable insights from these data. My belief is that artificial intelligence and machine learning in particular will play a crucial role in this process, especially now the data being collected are becoming increasingly complex.
Prediction models are not really part of traditional analytics tools. In my opinion, analytics tools are complementary to machine learning predictions. You can use analytics tools in the data pre-processing phase of a data mining project. In both supervised learning and unsupervised learning, analytics tools can be of use for first insights. However, if you want to perform machine learning predictions, with a defined target (hence supervised learning), you need to apply prediction or correlation models such as classifiers, neural networks, decision trees,… or other models.
Imagine your favorite dish. The taste, the smell… great, right? Now imagine that great dish with a pinch of salt. Initially, you wouldn’t even notice it’s there. Yet it enhances the flavor and therefore it makes all the difference in the world. Maarten Herthoge, Consultant & Team Lead Data Science at delaware makes an interesting comparison here. Read more about his view on Artificial Intelligence, and more specifically on his thoughts about the future of the digital workplace!
In 2017, Gartner put Artificial Intelligence and machine learning as its own trend(-in- the-making). This year, the trend has matured as we are now looking at how this technology is being integrated in day to day operations. “Although using AI correctly has the power to disrupt, it is not the holy grail where a general AI is able to learn any task that a human can learn and do. Instead, organizations should integrate & highly scoped machine learning solutions that target a specific task with algorithms chosen
that are optimized for that task” said David Cearley, VP and Gartner Fellow, and I can’t agree with him more. Let’s leave the overhyped, single-entity AI that can do anything to science fictions writers.
Take a look at the roadmap of many software vendors: you will find few that do not put AI on there. More and more apps will be backed by the power AI to the change the way we interact with IT and vice versa. Just like adding a pinch of salt, most users will find their experience now excels their expectations without realizing how much has changed.
The impact of AI can be subtle and perceived as a new feature. The new release of your sales dashboards enables you to have a quick look ahead using forecasting at the click of a button. Your video conferencing system is able to transcribe your spoken conversation for later reference or even translate in real-time. Instead of having people managing contractors on a production site, a bot takes care of handling the paper work and cameras track every moment to ensure floor safety.
“Apps will be backed by the power AI to the change the way we interact with IT and vice versa”
Instead of calling in sick, you send a text to a virtual agent who handles further communication with HR and sets your out-of-office while he’s at it.
If we look at businesses that are moving towards a more data-driven organization, AI can play the role of data pioneer. In every organization, people will try to formalize business logic one way or another, looking for new insights or automating aspects of their workplace. Consolidating these workflows with a classic business intelligence approach is often the first step organizations take but tends to leave users empty handed. Introducing AI in a business intelligence flow, leveraging computing power and machine learning techniques, one can move away from just looking at historical facts and gain new insights that previously went unnoticed or were too complex to grasp.
“Introducing AI in a business intelligence flow, can gain new insights that previously went unnoticed or were too complex”
This frees up people to focus on more complex, value-adding activities and to look at the future. As a side effect, building an AI powered digital workplace can encourage users to adopt this new data-driven way of working as their new toolset provides added value out-of- the-box and enables them to do more by default. Having a reinforced, user-friendly digital workplace is a strong incentive to move forward. This in turn creates a platform for broader digital transformation. In the digital workplace, AI will take up the role of catalyst, setting the pace for innovation and enabling people to exceed.
Today psychology has to rely mostly on self observation and self reflection as a starting point for intervention. AI helps to collect data over larger periods of time as a more objective basis for starting psychological growth and intervention. It is an external source of information that clarifies situations and brings awareness that sometimes isn’t there yet.
AI can also collect psychological information over large groups of people which helps to shed light on best practices.
The government is not immune to AI, democracy 2.0 will be based on Artificial Intelligence and full transparency, combined with the virtues of blockchain related technologies.
80% of Marketing Executives predict that AI will revolutionize their industry by 2020. They’re right. Artificial Intelligence will be disrupting your business, and is already doing so. Take a look at these successful AI implementations of today, some already deeply rooted in our day-to-day lives, others new and innovative.
Just look at your smartphone
Various customer-facing or customer-oriented AI applications are already very common. Think about virtual assistants such as Siri and Alexa, and applications such as predictive marketing or intelligent robots and demand forecasting are raising productivity and profits in supply chain and distribution centers.
Intelligent Marketing Assistant
Tour Operator Virgin Holidays is using AI-powered software from startup ‘Phrasee’ to automate and optimise the writing of its marketing email subject lines. A succesfull implementation, because the machines outperformed the ‘humans’ by a whopping 10% when it came to open rates!
AI replaced a team of 34
In Japan, insurance firm Fukoku Mutual Life Insurance replaced its team of 34 employees with IBM’s Watson Explorer AI. This artificial intelligence system can calculate insurance policy payouts by analyzing medical certificates, surgery, procedure data and hospital stays. The firm believes it will increase productivity by 30 percent and estimates to save around 140 million yen (roughly 1.052.800 euro). KA-CHING!
The NHS (National Health Service of the United Kingdom) is to trial an AI-powered chatbot to help relieve the 111 non-emergency helpline. The trial is running in North London, where around 1.2 million residents can opt for a chatbot rather than talking to a person on the helpline. The chatbot is created by Babylon Health and lets people enter their symptoms. The app then consults a large medical database and users receive a tailored response, based on the information they’ve provided. If the bot can’t help them, they’re put through to talk to a human helper.
Amazon predicts your purchases
Amazon’s AI algorithms have been in place for quite some time. They’re refined every year and are getting quite smart at predicting in what customers want to buy, based on their online behavior. While that’s no groundbreaking AI implementation, Amazon’s future plans are. They plan to ship products to customers before they even know that they need them. Creepy or genious, who can say.
Content writers – beware!
For those who thought copywriting would escape AI: think again. Tools like Wordsmith and Quill are already being used by Associated Press and Forbes to create news content. The applications can create own unique written content, you can’t tell if a ‘human’ wrote it, or AI. For now, these tools use templates and enter relevant data and keywords, but it won’t be long until they write articles from scratch.
Are you ready to start implementing Artificial Intelligence in your marketing,sales and overall business strategy? 2020 is less than two years away…
Find out more at our AI 4 Business Summit. Grab your tickets here.
Through this webinar Microsoft will introduce all the ways AI can help you transforming your business and we will see how data and analytics powered by the most trusted and intelligent cloud can help companies differentiate and stay ahead of competition, discussing 3 main topics