Can machines think? British mathematician Alan Turing asked that question in the 1950s, laying the foundation for artificial-intelligence (AI) research. Soon after, MIT opened its first Artificial Intelligence Laboratory. During the following decades, researchers attempted to build computer systems that could reason logically, understand natural language, make use of machine vision, and model the human brain. However, the AI inquiry always hit bottlenecks, usually due to lack of computing power, data, shortfalls in methods, or funding.
In the mid-2000s, breakthroughs in mathematics, computing power, big data, and cloud connectivity allowed the development of much deeper and more complex models. These advances occurred almost simultaneously: Like the burst of a supernova after a stellar collision, this convergence of technological advances caused chain reactions that propelled machine learning.
Google Maps is a simple example of how machine learning has entered everyday life. The geospatial street data that inform Google Maps’ models of road networks often contain errors or old information, but machine-learning algorithms can quickly correct errors using Google’s Street View imagery. If the original Maps data doesn’t show a stop sign at a particular intersection, but the Street View imagery does, the system can recognize the error and automatically correct it.
Data is the fuel that powers machine learning. The proliferation of cloud computing and the billions of sensors from the Industrial Internet of Things (IIoT) capture the data that makes machine learning possible. Effective machine-learning systems succeed based on the breadth, relevancy, and variety of the data they process.
Today, machine-learning systems can already recognize faces and detect cancer better than humans. But the evolution of machine-learning tools will happen in phases, continuing to streamline and bolster the process of making things and change the way people work. With machine learning, humans will become curators and facilitators in addition to creators: a symbiosis of human and machine. A world of expanded potential awaits, an age of unbridled creativity where there are no limits.
Here, learn more about AI, machine learning, today’s capabilities, and tomorrow’s possibilities.
Machine learning and artificial intelligence are often used interchangeably, but machine learning is actually a subset of AI. To understand the future of these technologies, it’s important to know what they are.
Traditional computers cannot learn or adapt to the information they receive. They simply hold it, doomed to repeat the same errors or fail to solve the same problems. Machine learning can teach computers to understand and interpret data and results without being explicitly programmed for specific outcomes. This allows the technology to make decisions based on past experience—in a nearly limitless number of scenarios and at incredible speeds.
At the core of its processes, machine learning takes large subsets of data and finds patterns in minutes or hours, work that would take humans days or even years. The computer can then apply those learnings and predict future outcomes. It might even be able to make recommendations for smarter, more efficient choices. But with machine learning, the computer does all of this without programming or instruction from humans: It harnesses the power of computing to “learn” on its own.
In simplest terms, machine learning is what you experience when you’re shopping online or researching a topic. Search engines or commerce sites observe what you’re looking at, clicking on, adding to your cart, or looking at for the longest time. The AI offers up more recommendations based on your past actions.
This is a type of machine learning called supervised learning. An algorithm collects data and feedback from a source and analyzes it to find relationships that lead to a specific outcome. That data can then be used with new data to predict better outcomes. As an example, this type of machine learning might help a company understand what customers like about specific products to highlight those attributes and entice customers into future purchases.
Unsupervised learning combs through sets of data and finds relationships or outcomes without being told explicitly what the outcome should be. This type of machine learning is helpful when you don’t know the answer you’re seeking. For example, if you’re collecting data from customers, you can use unsupervised learning to classify them into smaller clusters to which you can assign marketing or strategic communications.
Reinforcement learning is a simplified algorithm that aims to create the most reward with individual actions. In gaming, a reinforcement learning algorithm could learn a game with the understanding that winning is the desired outcome. It will seek out and select moves that are most likely to end in a win (a reinforcement of its choices), and over time, that algorithm could become capable of beating the computer systems that developed the game in the first place.
The human mind has many capabilities beyond what computers can do. But at the end of the day, humans can’t model everything. Humans make choices and assumptions based on the data available to them, for better or worse. But AI is going to radically change that and will also change the future of all industries and customer-business relationships. AI can simplify assumptions and decision-making—and can also change the way business is done, the way customers are earned, and the way solutions are found.
Creativity and innovation are often buried beneath the tedious tasks of everyday work. If AI can complete, eliminate, or simplify those tasks, that frees up designers to explore new ideas, for example. Rapid prototyping can happen at a faster pace. Machine learning can apply data to these prototypes faster than any real-world testing method. The result is a better product with a greater likelihood of success.
In the future, AI may transform one company into an entirely different type of company. Imagine a construction company that has done construction in a certain area of the market for decades. It has collected digital data over that time, so it has a data source that represents some of the best practices in that market. It may be able to publish that as a model, and could put it on a marketplace where another fledgling construction company could use it to transition more seamlessly. Now the original company is using its model license and has a digital revenue stream from data it has collected for years.
Most design problems involve so many interconnected factors, it would be impossible to foresee them all. With AI, it’s possible to see depth in data and complexity that was previously beyond human scope. Working with new ideas that can be developed when people have more time for ideation and innovation, you can explore possibilities stemming from micro choices, such as individual parts on a machine, to macro choices such as urban planning.
The companies that will lead the way are figuring out the secret sauce of combining human and machine capabilities. Digitizing isn’t just about connecting computers and wires to things. It’s about getting systems to actually work in their intended environments, making sure machines, a factory, a construction site, and media production pipelines are reaching their potential.
That puts companies in a powerful place to change what they do, to shift from just manufacturing or just being a construction company. They could be in a position to be a broader enabler of a new kind of digital ecosystem.
The Standish Group found that only 35% of projects are successful in current business models. That places a significant burden on the resources of a company, no matter the size. And the projects themselves have become more complex. In many industries, there are not enough people with the skills to tackle those projects—and an abundance of data and digital tools results in a fragmented workflow when a streamlined systems approach is needed.
The modernization of project management could help companies develop more successful workflows that reduce waste and inefficiencies. That, in the end, could lead to more projects in general, and more successful projects. In the simplest terms, projects mean business and business means growth. With AI, faster ideation can help teams dream, develop, and execute at a pace that far exceeds today’s capabilities.
The desired end result of using AI is that every person and project is more successful, and that’s especially important for customer outcomes. Machine learning can use data and results to help businesses close the gap between customer expectations and reality.
Companies can use AI-enabled tools to better predict customer needs based on previous purchases. They could send reminders to phones, emails, or machines when it’s time to reorder certain parts; using data from customers, these messages could be timed more precisely and deliver relevant services or secondary content.
AI-powered chatbots can provide real-time “interaction” between a customer and a business, helping navigate questions and expedite a resolution. If the problem exceeds the chatbot’s knowledge or learning, the customer can be funneled into a call log for personalized service. According to data from 2022, 68% of users enjoy the speed of response from chatbots.
Machine learning is a concept with profound implications for all industries. Its vast potential has only started to emerge, but currently, AI lacks the diversity of inputs that would reflect humanity’s creativity, curiosity, empathy, and other attributes. However combining human strengths with machine learning’s ability to directly interpret complex systems can yield transformational creative power.
AI is the field of science that propels computers to perform tasks that typically would require human intelligence and instruction. This science uses datasets to examine, explain, and predict outcomes for which humans don’t have the capability to solve. AI is the broadest term, often used to explain the other concepts.
Machine learning is a type of computer science that enables computers to “learn” from data sets on their own without programming or direct input from humans.
Deep learning, a type of machine learning, uses increasing layers of processing for understanding data. These layers of processing are inspired by the human brain’s natural data-processing structures and capabilities.
Solving some of the world’s toughest problems will require going beyond available traditional mechanisms and will require new approaches and new technologies.
A computer alone can’t solve climate change, for example. It doesn’t understand enough context. Nor can an individual human understand every element of a design that could influence the climate effects of, for example, a building. This is where highly creative people working with highly powerful software can realize the true potential of AI. But first, today’s roadblocks have to be removed. One of those roadblocks, complex software, often stands between the people with the ideas and the solutions they seek.
The design and make industry is still stuck in many old, file-based, incompatible processes and formats. Data is in disparate locations, making it unusable outside of specific programs or functions.
Take a traditional factory, with equipment running on old software. Interfacing with that equipment and putting CAD data into the factory presents a difficult problem because the software is old. If you send a design to a factory, that data has to be reinterpreted for the interfaces that speak to the equipment—and likely, only that specific equipment. It’s no wonder some factories, such as those operated by car manufacturers, take a year or more to change machining and production lines. Using AI-powered technologies that can speak across platforms and networks would allow companies to respond to market demands in entirely different ways, giving them a competitive edge.
Urbanization is another acute problem where machine learning could have a dramatic impact. To solve an urbanization issue in a city, there has to be enough real-world data about what’s going on. Computers can’t model what they’ve never encountered, and humans often ignore things that are too complex to process. The answer is to collect data, bring that into the systems, and make it part of a data flow. Traditionally, this type of work has been bogged down by the limitations of file-based software and the complexity of moving files from one place to another.
Cloud-based platforms eliminate those issues and clear the path for rapid innovation. Over the past few decades, companies using cloud-based platforms have been able to decompose data, break it into granular parts, and give the right data to the right person at the right time. This makes it much easier to build flexible workflows for every person and function in the company without the weight of inefficient software and interfaces.
By collecting and using data and cloud platform learnings for decades, Autodesk has developed solutions that propel customers to success. The demand for design and make services is growing exponentially, year over year. Customers need better AI-powered solutions for housing, smarter options for construction, and more scalable choices for production and rendering. And they need all of those things to be high quality and cost-efficient. Autodesk’s focus on implementing AI into customer needs centers on three areas:
Augmentation:The goal of AI is not to replace creativity—it’s not possible to replace the creativity that’s innate to human nature. But using computational techniques to deal with large amounts of data can help people visualize scenarios and outcomes beyond traditional human scope.
Automation:Creative work includes tedious, noncreative tasks that take a lot of time and effort. For example, a designer needs to publish a set of technical drawings to send to a factory. Using AI to produce those drawings frees up the designer’s time for more innovative and imaginative aspects of the process.
Analysis: Analysis is all about using data. Just collecting enormous amounts of data doesn’t make it useful. AI and machine learning can analyze that data, pick out the trends, pick up the signals, and reveal the insights that are really going to help.
A lot of these processes are already at work in the design and make industries and services. Forma uses sun-hour analysis to help maximize the light in a building and the comfort of its occupants. It can also use rapid noise analysis and rapid wind analysis to optimize for energy efficiency, sound pollution, and weather conditions.
Markup Import and Markup Assist make AutoCAD revisions and markups less manual and tedious. Functionality includes recognizing handwriting, converting it to text, and layering it over drawings. Then you can scale, move, or rotate the layers to share. Maya Assist uses AI-powered text prompts for enhanced animation, bringing incremental elements of design and visual effects into streamlined projects.
Despite the continued evolution of AI and machine learning in business today, some companies are reluctant to adopt these cloud-based platforms and AI for their customers. To make a big shift in their infrastructure, a step function in efficiency is needed. Companies won’t make a move if it only gains them 5% efficiency, or even 10%. But if a move brings a 500% or 1,000% jump, that’s worth the investment. And right now, AI is the only thing in the technology landscape that holds the promise to create that level of acceleration.
A classic example of this potential gain is in construction. Construction is everywhere, every day, in any city in the world. Construction often involves dozens of companies with collected data all over the place. It’s very hard in these environments to start solving complex problems because the data flow is so sticky; information is caught in one-off programs and files that don’t translate from one part of the process to another.
Cloud platforms can solve these problems. Not immediately, of course, but when data can flow through those platforms, insights are gathered, and eventually new workflows are produced—largely without human direction, allowing humans to focus on other elements of the work.
AI is poised to create a fundamental change in workforce dynamics and will be an increasingly useful technology to solve more problems in the world—and there is no shortage of problems.
The work of AI is emerging and still largely unknown, but it has the potential to create ecosystems of change that help people tackle tough problems, such as waste in the oceans and climate change, clearing the hurdles between today’s technology and tomorrow’s results.
Kimberly Holland is a lifestyle writer and editor based in Birmingham, AL. When not organizing her books by color, Holland enjoys toying with new kitchen gadgets and feeding her friends all her cooking experiments.
Emerging Tech
Emerging Tech
Emerging Tech
