AI as Key Exponential Technology in the Smart Technology Era2022-03-162022-03-16http://cortexgroup.ai/wp-content/uploads/2019/04/logo_header.pngCortex Grouphttps://cortexgroup.ai/wp-content/uploads/2022/03/ai_in_smarttechera.png200px200px
The start of the Democratizing AI Newsletter which focuses in the first edition on “Artificial Intelligence a Key Exponential Technology in the Smart Technology Era” coincides with the launch of BiCstreet‘s “AI World Series” Live event, which kicks off both virtually and in-person (limited) from 10 March 2022, where this theme, amongst others, will be discussed in more detail over a 10-week AI World Series programme. The event is an excellent opportunity for companies, startups, governments, organisations and white collar professionals all over the world, to understand why Artificial Intelligence is critical towards strategic growth for any department or genre. (To book your tickets to this global event click the link below and enter this Coupon Code to get 5% Discount: Enter Coupon Code: JACQUES001 (Purchase Tickets here: https://www.BiCflix.com; See the 10 Weekly Program here: https://www.BiCstreet.com)).
This article shares some text and audio extracts from Chapters 1-3 of Democratizing Artificial Intelligence to Benefit Everyone:Shaping a Better Future in the Smart Technology Era as it pertains to the following topics (which will also be discussed on 10 March 2022):
Introduction to AI to Benefit Everyone: Shaping a Better Future in the Smart Technology Era
The Smart Technology Era is Here
Some of AI’s Challenges and Rewards in the Smart Technology Era
AI as Key Exponential Technology in the Smart Technology Era: AI’s Transformative Impact on Our World
Some Brief Historical Highlights of AI
Demystifying AI and its Multifaceted Nature
Assisted, Augmented and Autonomous AI
The Intelligence of Things, Blockchain, and the Future of Computing
Our Responsibility in Directing AI
1. Introduction to Democratizing AI to Benefit Everyone
We live in tremendously exciting times where we already experience the disruptive and far-reaching impact of a smart technology revolution that seems to be on track to comprehensively change how we live, work, play, interact, and relate to one another. And it is happening fast, almost at break-neck speed, especially in relation to what we see in homo sapiens’ rear-view mirror. It is like being on a run-away train or imagine driving a car that keeps on accelerating where we only have control of the steering wheel. It is therefore also extremely dangerous times, where it is critical to think on our feet and make the right choices to not only “save” our lives and get control of the situation, but also shape a better future for all of humanity. That is a daunting task that requires visionary leadership, wisdom, innovative thinking across multiple disciplines, and comprehensive collective collaboration of all stakeholders and levels of society. Whereas some people feel helpless or choose to ignore or deny or are even blissfully unaware of the full extent of these developments, we cannot afford to rest on our laurels. As humans we experience time in a linear way and sometimes struggle to see the speed of technology change, driven by a combination of exponential technologies such as artificial intelligence (AI), the Internet of Things (IoT), autonomous vehicles, robotics, 3-D printing, nanotechnology, biotechnology, materials science, energy storage, distributed ledger technology and quantum computing. The possibilities of people connected by mobile devices, with unprecedented processing power, storage capacity, and access to knowledge, are not only unlimited, but can be multiplied and amplified by these exponential technologies. The more we instrument the world, the more data are being generated to feed our artificial intelligence algorithms, which in turn leads to intelligent automation, more powerful solutions, and significant technology disruption. With AI spearheading the smart technology revolution, we see fundamental changes happening simultaneously with new opportunities through infinite data, efficiencies through self-learning, and the ability to bring machine interaction closer to human interaction. The smart technology revolution could either pull the “bottom billion” out of poverty, speed up personalized education, help us to be healthier through personalized wellness and precision medicine, and transform dysfunctional institutions or it could entrench injustice and increase inequality (almost sixty percent of the people in the world own less than $10,000 — or roughly less than 2 percent of total wealth in the world).[i] The outcome will depend on how we manage the coming changes. It is clearly not in the best interest of humanity to have regions of the world left behind or get involved in a smart technology driven arms race. With wisdom, a deep understanding of reality and the systems that we have created, and proper use of our ever increasing smart technology toolbox, we have the opportunity to re-engineer and calibrate a more equitable, safe, and prosperous world where we not only see economic growth and other positive effects of intelligent automation, labor productivity enhancement, and innovation diffusion, but also decentralized smart technology-driven governance, meaningful work and relationships, people incentivized to make positive contributions to society, and a new type of sharing economy that benefits humanity as a whole.
2. The Smart Technology Era is Here
Before highlighting how AI is transforming the world, let us dig into the Smart Technology Era that we have entered. One of the amazing things about this time, which already has a significant impact on our lives, is that it has only just begun. Each adult and even each teenager has grown up in a relatively different time to the Era that has just begun. Of course, teenagers will experience less of an adjustment because they were not born in a time where dial-up internet or even no internet was a part of their realities. They have not experienced the changes that personal technology and digitization brought. They were simply born into a time where they already existed. They were born into a digital world where digital interactions and transactions are relatively commonplace in most parts of the developed world. This aptitude for digital life might make the changes that are coming simpler to adopt and consume, but this does not mean that these changes will be small, and it does not mean that they will be slow. In fact, they are coming fast. Faster than we have ever seen before. These smart technologies like nanotechnology, robots, IoT, drones and intelligent, cognitive machines, all used in conjunction and powered by the speed and processing abilities of advanced technology, cloud technology and huge amounts of online data available, are all coming together to give us the tools dreamed possible in Science Fiction. These tools, as will be described in detail throughout this book, are the enablers to transform the world. Of course, we cannot predict the future, but we can steer it intentionally in a direction that can help, protect, and enhance all life. To do this, however, we need an accurate understanding of where we are. We need to be informed, empowered to learn from history and equipped with how fast and with how much range these changes are coming. In the broader context, the human revolutions started with the cognitive revolution about 50-80,000 years ago with communications skills. This was followed by the agricultural revolution 12,000 years ago when foragers turned into farmers, humans gained mastery over animals and the rise of cities. With the industrial revolution coupled with the enlightenment and scientific method, homo sapiens started to gain mastery over the planet which led to ending the perpetual tyranny of famine, starvation, and extreme poverty. We are currently in an information revolution, where smart technologies such as AI will likely ensure more profound impacts than any of the other major human revolutions.
Let us begin by understanding history and using this as a way to help us place ourselves in the Transformation that has begun – the Smart Technology Transformation. Of course, this transformation has been given many names or is discussed in various contexts. Klaus Schwab, of the WEF, calls it IR4 or the 4th Industrial Revolution. Steve Case calls it the 3rd Internet Wave.[i]The Economic Singularity’s Calum Chace discusses this new transformation within the context of the Information Revolution where the dramatic growth in the capability of AI leads first to an economic singularity and then possibly a technological singularity.[ii] Yuval Harari, in Sapiens, Homo Deus and 21 Lessons for the 21st Century, discusses the twin revolutions in information technology and biotechnology within the context of the Scientific Revolution.[iii] Richard Baldwin calls it The Globotics Transformation.[iv] Erik Brynjolfsson and Andrew McAfee call it The 2nd Machine Age which started with the Digital and Information Revolution.[v] Each of the names given to the Era that has just begun are all in relation to the transformations or revolutions that preceded them. And each transformation or revolution in history has affected not only work and production, but agriculture, politics, economics, social climates, and the very way people go about their lives. We know these best as the 1st, 2nd and 3rd Industrial Revolutions throughout which globalization and automation have been creeping their way into our existence. We may have started off as local, rural, community driven creatures, more focused on our immediate survival and trading only what was immediately available in our proximity, but with each transformation we became more globalized, more connected and tasks, processes and services became increasingly more automated.
Let us take a brief look at the history of the revolutions or transformations over the last few centuries that have led us to where we are today:
● The First Transformation (the start of the Industrial Revolution) used technologies such as water and steam power to mechanize production. This began in the early 1700s where societies effectively switched from agriculture to industrial and from rural to urban. From 1712 onwards, we saw for example the age of primitive steam engines, textile manufacturing machines, and the canals, whereas from 1830 onwards the age of mobile steam engines and the railways. These technologies (along with electricity soon thereafter) unleashed the disruptive duo of automation and globalization which led to an economic transformation and economic and social upheavals. However, technology impulses launched new forms of automation long before they launched new forms of globalization (a century later) where we for example saw steamships and railroads dramatically reduced the cost of moving goods.
● The Second Transformation used electric power to create mass production. This began in the late 1800s, where we saw the age of steel and heavy engineering and the birth of the chemicals industry and from 1910 onwards the age of electricity, oil, mass production, cars, planes, and mass travel. During the second industrial revolution we witnessed how technology produced technology with a cluster bomb of innovations on the advanced economies where each explosion produced a chain reaction of innovation, rising productivity, and income growth.[vi]
● The Third used electronics and information technology to automate production and connect the world. The digital revolution turned analogue to digital and began around the 1950s and led to the Services Transformation which for all practical purposes started in the early 1970s. Whereas the past globalization and automation is mostly about making and shipping goods, globalization and automation in the Services Transformation is about processing and transmitting information (that is linked to the laws of physics that apply to electrons and photons, and not matter which is more restrained). We have effectively moved from things (which includes land to capital) to thoughts. Computers and other digital devices are doing for mental power (ability to use our brains to understand and shape our environments) what the steam engine and its descendants did for muscle power. Information and knowledge became increasingly important factors of production (alongside capital, labor, and raw materials) and acquired economic value in its own right. Although the industrial revolution is still ongoing, there was a shift in focus from industry to services, which mostly disrupted the manufacturing sector. Services became the mainstay of the overall economy, pushing manufacturing into second place and agriculture into third.
● And the Fourth Transformation, the Smart Technology Era, is building on the information and digital revolution to not only create smart automation, but new forms of globalization and robotics that taps into a wild combination of smart technologies where the distinction between the digital, physical, and biological realms are not clear. We can probably date this back to 2010s (although the exact times will likely be dictated by history books of the future). Richard Baldwin has coined these new forms of globalization and robotics into a new word called “globotics”, where tele-migrants and white-collar robots coming for the same jobs at the same time are driven by the same digital technologies. This globotics transformation applied to the services sector has an amazingly fast and unfair impact on societies, effectively disrupting the services sector in a significant way. The result is an upheaval, a so-called Globotics Upheaval, and a backlash for which we need a resolution.
It is interesting to note that some of today’s thought leaders do not see these transformations or revolutions as broken up into four. Calum Chace, author of The Economic Singularity believes that there have only been two real Revolutions in modern times.[vii] The first one that began in the 1700s and evolved as the domino effects saw one revelation or invention leading to another and then another. The second revolution was an Information Revolution that began when information and knowledge became such important parts of production and services, that it pushed manufacturing and agriculture into second and third places, respectively.[viii] That began with the age of computing – the digital revolution if you would like to call it that. On the other hand, Erik Brynjolfsson and Andrew McAfee in The Second Machine Age believes that we can look at the transformations in modern history as divided between the 1st Machine Age (also beginning with the steam engine in the early 1700s) and the Second Machine Age where computers, digital devices and the flux of information changed the very nature of our existence yet again.[ix] Jeremy Rifkin, in The Third Industrial Revolution, divides the world’s transformations into three.[x] The First Industrial Revolution, powered by the steam engine, took us from rural into urban. The Second Industrial Revolution changed the landscape of urban life with the telephone, fossil fuels and automobiles and the aspirational Third Industrial Revolution which is unfolding with the convergence of ultra-fast 5G communication technologies, a renewable energy internet and driverless mobility internet all connected by IoT. He is advocating for a 21st century smart digital infrastructure to give rise to a radical new sharing economy that is transforming how we manage, power, and move economic life.
There is much to learn from the patterns of previous transformations (call them industrial, economic or technological revolutions); from the developments that changed the course of the world and left those who feared, rejected or were geographically or economically sidelined in, what Richard Baldwin, calls “upheaval”.[xi] Whatever the defining nature of our revolutions, whether they have been seen to be primarily production-based, services-based and now, information-based – upheaval is one thing that they have in common. Their domino effects are predictable because of how these changes (in what may seem to affect only one industry in the beginning) spread into every part of life. This common factor of economic, social, or political upheaval is an important thing to note, as it reminds us that what is happening now has happened before in some way. But our urgency to understand this now is greater than it ever was before because our technological advances and powers mean that the speed of these changes, and their capacity to spread into every part of our existence means that there is no time not to be thinking about solutions, ways to manage these changes smoothly and the parts we will play in steering the information-powered Smart Technology Era. The commonly termed Industrial Revolutions have been a whole lot more than industrial. Yes, they have transformed industries, but they have also transformed economies, politics, societies, and individual lives.[xii] They have changed the world as we know it and presented an entirely different one. Sometimes the new world has taken centuries to shape, but as we become more developed, the time it takes for revolutions to affect mass change grows smaller.
3. Some of AI’s Challenges and Rewards in the Smart Technology Era
4. AI as Key Exponential Technology in the Smart Technology Era
Now that we have a broader perspective of the Smart Technology Era that we live and breathe, this chapter explores AI in more detail, how it impacts this world in a transformative way, where it all started and how it evolved over the past century and the many breakthroughs and historical highlights. I also unzip AI and its multifaceted nature and many subfields, its applications, its various flavors with respect to assisted, augmented, and autonomous intelligence, how the fusion of AI, IoT and blockchain are likely to impact our world in a significant way, and how the future of computing can make or break the AI revolution. Given what is at stake, our current civilization has a massive responsibility to direct AI in a visionary and wise fashion towards wholesome applications that maximizes the benefits to as many people as possible and life more broadly.
AI’s Transformative Impact on Our World
It has become quite evident during the last few years that Artificial Intelligence is one of the key exponential technologies in the Smart Technology Era, to such an extent that it has even been called the AI era which follows the Internet era of the past 25 years. AI’s transformative impact on our world is not only due to how it can unlock business and societal value from exponentially growing structured and unstructured data across an ever-increasing instrumented world, but also its pervasive nature – in some sense not too dissimilar to the role of electricity – and ability to be used in conjunction with other technologies to construct smarter and more powerful technologies enriched with AI. These smart technologies significantly strengthen our capabilities to solve problems, experiment, research, gain insights, discover knowledge, and engineer solutions in a multi-, trans-, and/or interdisciplinary fashion within a limitless universe of application use cases. Aside from what smart technologies are capable of doing for businesses by automating, predicting, classifying, deciding, comparing, recognizing, detecting and recommending, it can transform entire ecosystems and systems of production, management and governance on a local and global scale. Kai-fu Lee is confident that AI will soon enter the elite club of general purpose technologies (GPTs as described in The Second Machine Age) which “interrupt and accelerate the normal march of economic progress” with the steam engine, electricity, and information and communication technology the key ones.[i] While the steam engine and electricity ramped up productivity by deskilling the production of goods and services, information and communication technology have a “skill bias” in favor of highly skilled workers. AI, however, will not facilitate the deskilling of economic production and will instead take over the execution of tasks that can be optimized using data without necessarily requiring human interaction. Kai-fu Lee believes that the AI revolution will be on the same scale as the industrial revolution, but likely larger and faster. He discusses three catalysts that will accelerate AI adoption, the first being that AI solutions are just digital algorithms that are infinitely replicable and instantly distributable across the globe, the second is venture capital funding backing AI ventures, and third is having China also participating and on par with the West in both advancing and applying AI technology.[ii] PwC predicts that AI will add $15.7 trillion to the global economy by 2030 with about 70% of that gains likely to accrue in China and the US.[iii] In the report Artificial Intelligence is the Future of Growth, Accenture suggests that AI as a new factor of production has the potential to double annual economic growth rates by 2035 through intelligent automation, labor and capital augmentation, and a driver of innovation.[iv]
Within the Smart Technology Era some of the core technologies that are changing and will continue to change our world include networks and sensors that instrument the world, infinite computing, AI that provides expertise on demand, robotics which act as our new workforce, genomics, and synthetic biology. According to a 2020 Forbes article, the top technology trends that will likely define this decade also includes AI and machine learning applications (of which many I’ll discuss further in this book) such as natural language processing, voice interfaces and chatbots, computer vision and facial recognition, robotic process automation, mass personalization and micro-moments, big data and augmented analytics, as well as other complementary smart technologies and applications such as IoT, robots and collaborative bots, autonomous vehicles, drones and unmanned aerial vehicles, wearables and augmented humans, intelligent spaces and smart places, blockchains and distributed ledgers, cloud and edge computing, digitally extended realities, digital twins, 5G of cellular network technology, genomics and gene editing, machine co-creativity and augmented design, digital platforms, cybersecurity and resilience, 3D and 4D printing and additive manufacturing, nanotechnology and materials science, and quantum computing.[v] The fusion of many of these exponential technologies along with the speed of technology change will likely lead to an exponential growth trajectory in the Smart Technology Era. AI is the key technology in this change, without which the extents and uses of these other technologies would be limited to less cognitive behaviors and outcomes. For example, often with supporting IoT, robotics, blockchain, nanotechnology, biotechnology, smartphones and internet, AI is already being used in detecting water, air and crop health, diagnosing diseases, predicting when maintenance is needed to avoid expensive or disastrous breakages or outages, analyzing electricity, gas and water wastage, advising on the best places to plant crops for greatest yield, providing psychiatric and medical diagnoses and advice, and allowing people of different languages to communicate with each other in text, voice or video (even across slang, accents, cultural and regional syntax, meaning variations, personal patterns, and figures of speech). AI has great power to be used to positively transform people’s experiences of the world, alleviate poverty, offer access to jobs and education, ensure all areas and communities have access to food and clean water, and more efficiently and effectively use what is already available and what has already been produced (circularity). The nature of smart technologies allows us to invent what was previously unimaginable, solve problems that we thought were just part of life and bring resources and vital services to those previously excluded.
AI is made even more powerful and omnipresent when combined with other smart technologies. In many cases, the use of AI is but one part of the fascinating inventions, which are really collaborations and integrations of different technologies. IoT, as the ability to power any device with the internet to keep it constantly connected and pulling data and statistics, can communicate with software halfway across the world. This software uses machine learning algorithms to predict maintenance, optimize efficiency, conclude outcomes, or gain valuable insights. It sends all this back to the device which then changes its track, alerts for maintenance, or redirects its power flow or focus. Already, artificial intelligence combined with other smart technologies, like IoT, is all around us, from self-driving cars and drones to virtual assistants and software that translate or invest. Impressive progress has been made in AI in recent years, driven by exponential increases in computing power and by the availability of vast amounts of data, from software used to discover new drugs to algorithms used to predict our cultural interests. Smart watches empowered by IoT and AI, can know when we are going to have a stroke or heart attack, or going into diabetic shock, can let us know when we are sick, even what the diagnoses is and tell us what we need to do, what we can take and where the nearest pharmacy that stocks what we need is.
In business, AI also has the power to enhance, automate, improve, predict, personalize, and optimize products, services, and processes. Together with its smart technology counterparts, AI is transforming businesses: business models, hierarchies, strategies, workflows, services, and products. To be able to achieve this, we must know what we are facing. We need to know the threats and risks so that we can ensure they do not occur. As much as we imagine the wonderful possibilities ahead of us, the truth is we have no idea what our future looks like – even less of an idea than we had during the previous industrial revolutions.[vi] We do know that the Information Revolution leading into the Smart Technology Era is the fastest-ever period of technological innovation. Whereas human brain power played a key role in solving the problems of the 19th century and software programs running on computers helped to address 20th century obstacles, AI and its applications are a key part of the 21st century solution stack to tackle today’s challenges. Living in this hyperconnected age, with more knowledge, more access to knowledge and the ability to make more sense of our knowledge, our lives are being continuously transformed and disrupted.[vii] Almost every industry and government is being transformed and disrupted by the knowledge driven, hyperconnected Smart Technology Era. Keeping up with these changes is where we find challenges. It is easier for economically stable countries and individuals who are exposed to the latest developments and have digital skills and education. It proves more difficult in countries and communities where the world’s advancement is sparse. The World Economic Forum (WEF) and Accenture’s warning that we could be looking at job losses as high as 2 billion by the year 2030, is a real fear in a world that is split between those who are part of the digital socio-economy and those who are not.[viii] If technology is the future, then only those who are empowered to use this technology are secured a spot in this future. We are only at the beginning of smart technology transformation, and already the lives of the tech-enabled citizen and the non-tech-enabled citizen are so far apart that, on a surface level, they are barely relatable. The non-tech-enabled citizen who, without a smartphone, computer or internet is not only living in a vastly different way, but is exposed to a fraction of the information, education, goods, trends, processes, and services than the tech-enabled citizen. Already, whether we are exposed to technology, and how much technology, is largely due to socio-economic status. The further danger is that technology has the potential to solidify the world’s socio-economic status and drive them further apart. Therefore, it has never been more important for every person to be educated on and included in the changes of the Smart Technology Era. We are on a tightrope, trying to balance the positive, transformational possibilities that AI brings, with the negative, socio-economic decline that, without immediate intervention, it has the potential to bring too. The wealth and opportunities of the world are already in the hands of the few. Smart technology and a digital economy could make this exponentially worse or exponentially better.
5. Brief Historical Highlights of Artificial Intelligence
6. Demystifying Artificial Intelligence and its Multifaceted Nature
7. Assisted, Augmented, and Autonomous Artificial Intelligence
8. The Intelligence of Things, Blockchain and the Future of Computing
9. Our Responsibility in Directing AI
Even though AI mimics part of the human process of reasoning, adapting, and processing information, it is not, at its core, about replacing humans with machines. It is about harnessing the combined strengths of both humans and machines to process environments and solve complex problems from constantly changing factors and arising information. As we move more deeply away from the “programmable era” of computers – where we have explicitly told computers what to do, we move into a space where we give computers the tools to tell themselves what to do or how to do it. And to become better and better at it, the more they learn. This is cognitive computing. It is the ability to mimic the human brain, to learn and to understand within the context that humans provide and, in this, be more of an assistant than a tool. It understands (by sensing and interacting with data), reasons (generating hypotheses and recommendations) and learns (what the lessons from masses of data are). It can take knowledge from different sources, bring it together, process and understand it without human involvement in every step. Some instances of machine intelligence are very specific to solving only certain problems or performing highly specific tasks. Part of this has to do with the data we have available to us, how that data is integrated and our own capacity to help the machines verify their learning pathways and conclusions so that they may continue learning, even from their mistakes. Because data, and large amounts of it are so important for any learning to take place, technology business giants such as Google, Apple, Facebook, Amazon, Microsoft, IBM, Baidu, and Alibaba are capitalizing on their data and are vying for ‘AI throne’. Before digital technology and massive processing power existed, data existed in separate entities, not accessible or able to pass through one place or virtually accessible. With digital technology, our ability to capture and store data in place, and increases in processing power which allow us to do this with massive amounts of data, we now have rapid access to all available structured and unstructured big data in one place physically or virtually. Big data refers to the large, diverse sets of information coming from multiple sources, arriving in multiple formats, and growing at ever-increasing rates. It encompasses the volume of information, the velocity or speed at which it is created and collected, and the variety or scope of the data points being covered. Big data often relies on Application Programming Interface (API) integration between different applications (data from different places, now able to exist in one place). Most applications or software components typically have an API, which is its computing interface that defines how other applications, systems or software components can use its features, functionality, or data via a set of routines, functions, protocols, or procedures that specify the kinds of calls or requests that can be made, how to make them, the conventions to follow and the data formats that should be used. Just think of all the different applications or programmes you use daily. The ability for these to speak to each other, not only makes your life easier on the surface but allows data from these different places to communicate, share and interact with each other in the background. We are indeed living in the API Economy. The combination of data and machine intelligence is also what powers AI-as-a-Service cloud-based solutions, frameworks, development environments, platforms, and applications. An AI-as-a-service solution can for example provide the results of trained ML models in the form of inferences. ML algorithms rely on data, teaching and knowing which results were correct in order to infer, in future, the conclusions based on more and more learning and more and more data. These inferences are the results we see when we are fed certain ads on Google or Facebook. The ML algorithm has inferred the content that would be of interest to us based on learning our behavior and interests. One of the key application areas of AI is data mining which involves the process of discovering patterns in large data sets that makes data usable, and less random. It looks for similarities, differences, relationships, and anomalies in the data to learn and reach conclusions. Data mining occurs in many layers, and as machines learn, the layers become deeper, sometimes finding patterns and discoveries that humans themselves may never find. To mine data, data often needs to be cleaned or prepared to deal with raw data that often exists either in duplicates, in incorrect data entries or in contradictory data entries. Cleaning the data looks for these instances to try and remove or correct them so that the data can be used. Now that we have massive amounts of data in one place, information engineering deals with the distribution, analysis and use of this information. In ML, the aim is to generate, understand and use the data in a way that supports learning and inferences. Often data exists as a small sample of the population and thus is not a true reflection of the full picture. On top of this, because data is reliant on humans to add, add value to, clean and mine, the data reflects what these humans deem important, the conclusions they are trying to find, or the value inherent to themselves. It is important to understand how easily bias occurs before training any ML model.
As the potential applications for Artificial Intelligence is limitless, it will have a transformational impact on all industries. Understanding AI and its applications are vital to our lives, livelihoods, and the future we are creating with every present action or inaction. Artificial intelligence is not some distant future, nor is it something we can escape. It is here. For now, it is a black box for most people. But “when a new technology is as pervasive and game changing as machine learning, it’s not wise to let it remain a black box.”[i] For how do we affect or direct what we do not know about or understand? How can we be a part of what we do not understand? More importantly, how can we control it if we do not understand it?[ii] We may think it necessary to control it, but just leave it up to the experts to understand it. But AI affects every part of life. It affects us, directly, and we should have a say in the things that affect us directly. Because if what we do not understand (and therefore cannot steer or control) are so much a part of the world around us and the tools we have as human beings to live, do business, to provide, to gain services and have insights into things, processes and each other, are we not intentionally removing ourselves from the core of the world as it is. It might not fit into our perception of the world as we know it. It might fit into our perception of the world as we wish it to be. It may not fit into what feels safe and known and comfortable, but it is still the reality we have and by denying it we are only excluding ourselves from being a true part of it. It is time for that to change. Some of the key reasons to prioritize learning about AI includes (1) adapting to the speed of AI implementations; (2) the fact that every major technology company is prioritizing AI; (3) companies that are first in deploying AI-driven solutions have competitive advantages over those that do not; (4) most countries are implementing new laws and regulations regarding smart technology that will likely affect everyone; (5) to ensure ethical, responsible implementation of AI applications; (6) more benefits and opportunities are likely for productive members of society that work together with smart technologies; (7) ensuring better collaboration between private and public sectors; (8) there is a shortage of knowledge workers such as data scientists, machine learning experts, and other technical professionals who can build AI solutions and services; (9) and the potential impacts on society.[iii]
The sheer intelligence of the systems makes them feel less like technology and more like a natural, human interaction where little effort or learning on our parts is required. This is contrary to the introduction of technology and digitization where the onus was on us to learn, train, adjust and come to terms with how to use completely new and sometimes complicated systems and processes. Even learning how to use a keyboard is an example of the effort humans had to put in to start using computers over traditional pen and paper. Still today, we are seeing low adoption rates in organisations that have tried to replace traditional ways of working with digital tools. While these tools are becoming simpler to use, without AI, there is still a large amount of learning, training, and adopting involved, which people sometimes feel is simply not worth the effort. Cognitive computing changes all of this. It makes the complexity of technology “disappear” – where using technology no longer feels like using technology. In fact, it will feel easy, natural, and intuitive for us to interact with smart devices. We are already seeing the beginnings of this with Google Assistant, Siri, Alexa, and Cortana. As we go deeper into untapping AI’s capacity and power, intelligence will be so infused into systems and devices that may not even be aware that they are using technology. There will be no effort. We are on a steady path towards a hybrid future full of diversity where multiple AIs are interconnected with each other and even, perhaps one day, within us. Whilst we are years away from seeing the effects of human biology infused with AI (e.g., in the form of chips, contact lenses, and so on), intelligent chairs, buildings, glasses and cars are here and will soon be naturally interacting with each other and with us.
We may think of algorithms in terms of AI only, but our brains use algorithms to see, hear, feel, learn, and understand. In fact, our brain uses the same algorithm to do all of this, and depending on the task, there are special parts of the brain that do the work and receive and send signals to either other parts of the brain or other parts of the body. However, if one part of the brain was damaged, for example, it is possible to direct those signals to another part of the brain, which would then become the new home of those signals and their subsequent tasks or results. This was demonstrated when a group of MIT students swapped around the eyes and the part of the brain responsible for sight (visual cortex), with the ears and the part of the brain responsible for the hearing auditory cortex. The result was that the Ferret was able to learn how to see and hear again.[iv] This is because, whilst firing different neurons in different directions and between different functions or parts of the body and their ‘home’ in the brain, the same algorithm is used. The intelligence is the same, it is only the location that is different and depending on the function or complexity requires a different number of neural connections to effectively work. In Yuval Harari’s Homo Deus, he emphasises that an algorithm is not a particular calculation, but the method followed when making the calculation. It is a methodical set of steps that can be used to make calculations, resolve problems, or reach decisions.[v] Humans do this all the time – each time we are faced with a decision, it is our own algorithms that decide how we will react. For example, if you see someone pushing in line, and that is met with a belief that pushing in lines is disrespectful, while you feel your purpose in life is to show people how to be more respectful, you are going to point out that the person pushed in line. Your own internal algorithm has led you to do this. Let us say the last time you politely showed someone how rude they were, you were punched in the face and left awfully embarrassed. This is now part of the data that your algorithm is processing. So, even by deciding not to point this out, no matter how much you want to, you are still functioning in the framework of an algorithm.
Your mind’s algorithms are constantly at work. We use our internal algorithms when we are choosing the best way forward as well. We have all been in situations where we must decide which choice will lead us to our desired outcome. Sometimes, this happens instantly and subconsciously, and sometimes we have time to weigh up the potential risks versus benefits. In both conscious and subconscious matters, we are analyzing the information before us (current data), past lessons (past data), personal beliefs and values (priorities for desired outcome) and things we would like to avoid (risks). We also do this knowing that some things have higher priority than others. Only the most mindful of you are aware that this process or some like it even occurs. You have most likely gone much of your life without even being aware that it exists. AI (machine learning, deep learning and everything that results from these) works in very much the same way, only instead of inhabiting the brains, we are creating them. We are creating the algorithms that tell the systems what to value, what to avoid, what to favor. We are ensuring that they learn from new situations and incorporate their learnings into future decisions. We tell them what to look at, how to weigh what they are looking at and how to know if they do not know the answer. The machine intelligence we are creating is our own intelligence. We are figuring out how our minds process data and make decisions so that we can guide machines to follow the same pathways and learn the same ways we do. Jeff Hawkins believes that we will not be able to create fully intelligent machines until we understand how the human brain works.[vi] This has been a question for psychologists, philosophers, biologists, neuroscientists, and neurosurgeons for many years. Everything the machines do is a result of humans. Everything they cannot do is a result of humans. Machine intelligence without human intelligence simply would not exist. It is important to remember that while these machines appear to be thinking for themselves, they are following a strict set of instructions and steps. Whatever they learn and however more precise or intelligent they become, they are still following the algorithm, or the automated learning process created by humans.
Our responsibility to direct AI towards favoring the good of life (all life, not just human life) is vital. To do this, it is important to understand our own algorithms and what makes them different. Each of our personalities and emotional responses, as Yuval Harari points out, is a result of our algorithms (however much more complex). Whilst certain steps in our algorithms might be the same as our collective past data (evolutionary needs and developments), personal past data (history), current data (present situation), outcomes (values and goals) and weighting (priorities) differ.[vii] They will even differ in different situations. For example, our values or goals on a work project may differ from our values and goals for our careers which will also differ from our values and goals in social situations. Making decisions in each of these situations will use a variety of different conscious and subconscious algorithms. If we have been taught to value privacy, autonomy, and self-actualization over harmony and collective good (a common value contrast between West and East), these will appear in the algorithms that make decisions.[viii] Similarly, if we grow up in a community, family or greater society that sees black males as more dangerous or women as less capable, these too appear in our thought and decision-making pathways. The danger with machine intelligence is humans are creating it. The algorithms we create are therefore just as susceptible to the prejudices, values, knowledge, and biases that we have.[ix] The only difference is that the effects are on a much larger scale. Data bias is a huge problem for the world of machine intelligence and the human element in creating AI is only part of why biases exist. There is also the fact that not everyone’s views, needs and priorities are reflected in the data that exists. Take Social Media for example. 2019 statistics show that Facebook users are at 2.4 billion, YouTube is at 1.9 billion, Instagram is at 1 billion, and Reddit and Twitter are both at 330 million.[x] If we are to take the information, sentiments and analyses we get from Social Media and use that to make decisions for the world, we are using data from not even half the world’s population to direct further decisions and recommendations. It is ground for a disaster of biases. In later chapters, I will provide insights into how data bias can be avoided, checked, and accounted for. For now, the most important thing to note is that machines can only be as smart as we are. We need to stop seeing them as separate and start seeing them as the tools we create and use to achieve our own goals.
I will conclude this chapter with some thoughts and sentiments from Demis Hassabis, a co-founder of Google Deepmind, that I also identify with and could materialize if we act with wisdom and take full responsibility in directing AI in ways that benefit as many people as possible. He believes that AI “could usher in a new renaissance of discovery, acting as a multiplier for human ingenuity, opening up entirely new areas of inquiry and spurring humanity to realize its full potential” and that it is likely going to be the “most important technology ever invented”.[xi] Demis further states that “by deepening our capacity to ask how and why, AI will advance the frontiers of knowledge and unlock whole new avenues of scientific discovery, improving the lives of billions of people” and that AI can help us “build radically new and improved ways of life” and through our curiosity, the scientific method and our use of AI to not only solve society’s greatest challenges today, but to understand ourselves and make sense of the universe around us.[xii]
Background LinkedIn Articles on Democratizing Artificial Intelligence