For a concept so crucial to the future of the world—and the world of business—artificial intelligence (AI) remains misunderstood by many. Mention AI and it still might call to mind images of Terminator, scheming computers, or robot armies.
Of course, if you’re here, you likely have some idea of AI’s importance to real-life innovation and future growth. In this article, part one of our three-part series on AI and innovation, we’ll clarify that importance and offer insights you can bring back to your business.
Explore this guide to AI and you’ll find:
- A basic definition of AI, machine learning, and deep learning
- An overview of AI’s origins
- Analysis of how AI can impact all parts of a business
- A look at the importance of ethics in AI
What is Artificial Intelligence?
The Encyclopedia Britannica defines artificial intelligence, or AI, as “the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings.”
While that sounds simple enough, the broad nature of such a definition is actually the subject of a long-running and heated debate within the field of computer science. After all, what exactly amounts to “intelligence” is the subject of debate among experts in the field.
Still, the crux of AI is there in that definition. In the simplest terms, AI is about machines that mimic human intelligence. And they do so in order to perform specific tasks and improve themselves (or “learn”) based on the information they gather.
Machine Learning and Deep Learning
While we’re establishing the basics, there are two terms we should define before moving further: machine learning and deep learning. As they’ve both become buzzwords in recent years, it may be easy to confuse them for one another—or lose sight of their place within the field of AI.
Image source: IBM
As the image above illustrates, machine learning and deep learning are actually branches of the wider topic of artificial intelligence.
Machine learning gives computers the ability to learn without being explicitly programmed to do so. Instead, computers imitate the way that humans learn—they analyze historical data inputs to predict output. As they learn more through repetition, the computers can predict these outcomes more accurately.
Deep learning, then, is a subfield within machine learning. It’s considered an evolution of machine learning. Deep learning applications consist of many-layered “neural networks,” complex sets of connected algorithms that mimic the activity of the human brain. Deep learning applications are able to learn on their own and process even more complex sets of data on their own.
Image source: IBM
Machine learning and deep learning have been used to fuel breakthroughs such as:
- Recommendation engines: Analyzing user behaviors, algorithms offer recommendations on streaming platforms and online retailers.
- Stock trading: Automated trading platforms make high-volume trades based on analyses of past market dynamics.
- Speech and image recognition: Computers are fed natural language and image inputs, improving their abilities to translate speech to text and identify content within images.
The History of AI
To understand the state of AI today and its future, it can help to know where the concept came from.
Most histories of AI trace the concept to mathematician Alan Turing’s landmark 1950 paper, “Computing Machinery and Intelligence.” Turing, already famous for his role cracking encrypted Nazi codes during World War II, gave birth to the field of artificial intelligence by posing a simple but profound question: “Can machines think?”
Turing also established a test, “The Imitation Game,” also known simply as the Turing Test, which ran computers through a series of reasoning puzzles to determine whether a machine could actually be considered capable of thinking.
The validity of this test has been debated by computer scientists ever since. However, Turing’s initial question still drives the dynamic field of AI today, as corporates and scientists pursue solutions that, at their core, are driven by machines that attempt to think and act like humans.
Two Types of AI
As they are developed, these solutions fall within two general categories, weak AI and strong AI.
Weak AI refers to solutions that are trained to perform specific, narrow tasks. (Weak AI is sometimes also aptly referred to as “narrow AI.”)
The vast majority of today’s AI applications fall within this category: question-answering solutions and voice-based assistants like IBM Watson, Apple’s Siri, and Amazon’s Alex are all Weak AI systems that are trained to analyze and answer questions posed in natural language. Likewise, autonomous, self-driving vehicles are narrow solutions focused on the task of driving without human input or control. Weak AI capabilities are typically powered by machine learning and deep learning.
Strong AI, also referred to as artificial general intelligence (AGI), refers to a form of AI where a machine possesses a wide-ranging intelligence similar to a human being’s. Rather than focusing on a narrow set of tasks, the AI would be able to independently approach, analyze, and solve any problem. According to IBM, Strong AI is still largely theoretical—any potential solutions remain under development.
The Growing Importance of AI Today
While the long-sought dream of Strong AI solutions are still being developed, AI is already delivering incredible innovations in the present, as more and more businesses leverage AI-powered capabilities.
According to the 2021 edition of McKinsey’s annual Global Survey on AI, 56% of respondents reported AI adoption in at least one function, up from 50% in 2020. Meanwhile, a PwC survey found that 86% feel that AI is becoming a mainstream technology at their organization.
Underlying this accelerated adoption is an increased recognition in AI’s ability to drive innovations across a range of business functions. A report from The AI Journal found that, in a post-COVID-19 world:
- 74% of leaders expected AI to make business processes more efficient
- 55% believe AI will help create new business models
- 54% expect AI to enable the creation of new products and services.
Trends in AI Capabilities
More and more decision-makers are recognizing the potential of AI. And as our increasingly digital world generates more and more data, in a sense we only create more opportunities to harness AI.
After all, AI’s greatest strength is in processing and analyzing data. AI is able to identify types of data, discover connections between different points and sets of data, and offer insights and visualizations of those connections. From IT to HR, the sales team to the supply chain, this processing power holds almost limitless potential for today’s businesses. See how some are already using it below.
Today’s IT teams are tasked with analyzing unprecedented amounts of data across a range of systems, applications, and tools. Meanwhile, these teams are often equally widely spread—and natural silos can make efficient information sharing difficult. And when a single outage can trigger thousands of logs and events—not to mention millions in losses—rapid response is essential.
That’s why AIOps, or Artificial Intelligence for IT Operations, has steadily gained traction in recent years. With the ability to replace or supplement a swath of manual process, AIOps can rapidly detect threats and anomalies, identify root causes of issues, and help teams get critical functions back online. Plus, machine learning-powered predictive maintenance can proactively identify potential issues before they become costly problems.
Watch: Dig Deeper into AIOps
Supply Chain Management
Of course, keeping business processes and customer experiences up and running is far from the only benefit AI can deliver. It’s also helping organizations process data to streamline and optimize these processes and product experiences.
Complex supply chains hold huge potential for loss and risk. They also generate incredible amounts of data. AI-powered analytics can aid companies with “advanced scenario modeling.” Creating a virtual replica of the supply chain in the cloud, AI scenario modeling analyzes end-to-end supply chain touchpoints in unprecedented depth. Modeling data is parsed to identify weaknesses and risks, optimize processes, and address inventory challenges.
Marketing and Sales
Image source: Smart Insights
Marketing and sales departments can harness AI’s powers of data analysis throughout the customer lifecycle. When the vast majority of today’s customers expect personalized experiences, AI algorithms help deliver.
Crunching large data sets and drawing behavioral insights, AI can ensure that content is adjusted to suit the user it targets. (Innovations in language processing from companies like Copysmith and Copy.ai are also helping marketers generate that content!) Plus, media budgets and bidding strategies can be optimized with the help of predictive analytics. And app experiences for end-users can be tailored based on individuals’ tendencies and preferences.
One of the more prominent examples of AI’s end-to-end potential in these areas, Netflix uses AI to guide series development decisions, place advertisements, provide users with personalized recommendations, and test app layouts and thumbnails for optimized engagement.
An Augmented Workforce
While the growth of AI and automation has driven fears of job elimination, it’s also creating jobs and helping job seekers find them. In fact, a World Economic Forum report estimated that AI would drive the creation of 97 million new jobs by 2025. Moreover, it’s clear that much of AI’s power lies in its ability to augment organizations’ existing talent.
Beyond sales and marketing, HR departments are harnessing AI to speed up recruiting. A global enterprise like Unilever has to process as many as 1.8 million job applications in a given year. Partnering with AI recruitment specialists Pymetrics, they created a multi-step process to assess candidates and narrow their applicant pool. In the end, HR chief Leena Nair estimated that the endeavor saved about 70,000 hours of work.
However, highly specialized workers are getting help. For instance, in the healthcare industry, AI isn’t just helping researchers cut trial costs and supporting patient outcomes with improved diagnostics. It’s also freeing up providers’ time to focus on more important tasks—tasks only a human can perform. Soon, healthcare robots like Mabu from Catalia Health and Beomni may be assisting teams by monitoring patients and performing diagnostic tests.
Watch: See How AI is Changing Healthcare
The Ethics of AI
Of course, while AI’s potential is clear, certain ethical issues associated with its rise demand continued attention and consideration.
AI requires the widespread collection, analysis, and use of huge sums of data. As such, the companies who develop and utilize AI capabilities are subject to reputational, regulatory, and legal risks related to the responsible sourcing and handling of that data.
But it’s not just data that poses ethical risks. The deployment of algorithms can ultimately incorporate and amplify human biases. As a team of McKinsey analysts write in the Harvard Business Review, “AI systems learn to make decisions based on training data, which can include biased human decisions, or reflect historical or social inequities, even if sensitive variables such as gender, race, or sexual orientation are removed.”
To counter the potential for bias to creep into AI, they recommend business leaders take steps that include:
- Keeping up-to-date on research with resources from groups like the Partnership on AI and the Alan Turing Institute’s Fairness, Transparency, Privacy group
- Establishing responsible processes that can mitigate bias. Google AI provides a series of recommended AI practices.
- Diversifying the field of AI with investments in diverse talent, education, and advocates of advancement like AI4ALL.
By taking the right steps and making investments to support the advancement of an ethical AI, corporates can help maximize the benefits AI technology offers to both businesses and the wider world.
Watch: Learn More About Keeping Bias Out of AI
Conclusion: Building the Future of AI
With an understanding of the awesome potential we’ve outlined, it’s no wonder why investors are putting more money into AI startups than ever before. According to CB Insights, global AI startups raised more than $50 billion in 2021—an amazing 55% increase from 2020—across over 2,000 deals.
These startups will be among those that fuel growth and change in an array of industries, from autonomous fleets in the transportation sector to those healthcare innovations we touched on above.
AI will also help drive entirely new innovations and whole industries, like the Metaverse, which we detailed in a recent guide. Microsoft made its play in this emerging space, with its recent acquisition of Activision Blizzard. And Facebook underscored the crucial role AI will play in making the Metaverse a reality.
With the insights of this guide, you can take your first steps, along with them, toward shaping the future of AI and innovation.
About Runway Innovation Hub
Runway is a Silicon Valley-based innovation company accelerating the success of global innovators and entrepreneurs. We help corporations through results-focused innovation consulting and power the growth of startups through acceleration programs, mentorship and coworking services. Since 2013, we have has accelerated the innovation efforts of 40+ global companies like Fujitsu, Lenovo and IBM through customized innovation strategy services. Learn more about how we can help you do the same here.