This is the third in a series of opinion pieces from campus leaders on the role Michigan Tech innovators will play in defining the world’s emerging needs.
From mainframes to personal computers to today’s mobile devices, the way we interact with computing has changed profoundly over the course of a generation. Each of these transitions represented a major inflection point in the ubiquity and centrality of computing – and it’s happening again today. ChatGPT has ushered in a whole new future, spurring artificial intelligence arms races between nations, businesses, and researchers. It is no exaggeration to say that all aspects of society and industry will be impacted by advancements in computing and AI. This is particularly true in prestigious sectors like healthcare, mobility, finance and entertainment, but also in less flashy areas like crafts, manufacturing and retail. .
These advances will undoubtedly lead to increased productivity and usability, as well as more personalized services, but will also introduce new challenges. For example, smart policing using data science and AI to identify high-risk areas in which to deploy police services is a very attractive idea at first glance; However, too often, these tools reinforce prejudices and calcify generations-old socio-economic inequalities, causing as much damage as good. Excitement about these technologies and their rapid adoption will need to be tempered by deeper analysis and mitigation of unintended consequences.
In the following paragraphs, we present three pressing questions regarding the future impacts of AI. The list is not intended to be exhaustive. We hope this illustrates that while computational expertise is the foundation of AI algorithms, a broader body of knowledge is required to apply them responsibly and effectively. Building on this, Michigan Tech can position itself as a leader in applying AI to other fields.
How will AI-driven design reshape product development?
AI can be creative, meaning it can discover solutions that humans haven’t explored. This has already happened with complex games like chess and Go, and it is also happening in the field of engineering, where analysis and design tools with AI are rapidly being integrated into the field.
As a broader set of solutions are presented and AI undoubtedly becomes an even more valuable tool for evaluating and optimizing multiple simultaneous object functions, the expertise and insight of the engineer becomes increasingly valuable. crucial importance. Engineers must have a deep understanding of the domain while working in cross-functional teams. They must also understand the quality and limitations of computer models and be able to assess the impact of changes not only on performance, but also on all requirements throughout the product lifecycle.
AI models are inherently non-deterministic. So when the engineer uses them to create a design, there are many possibilities available to him. It is then the engineer’s responsibility to critically evaluate and analyze the designs, ruling out problematic options and refining those that show promise. Final choices must be thoroughly reviewed and validated effectively and efficiently through testing under real-world conditions.
Michigan Tech, known for its emphasis on hands-on learning, is uniquely positioned to integrate AI into engineering design.
We will need to ask ourselves questions such as: How can we ensure transparency throughout product development to ensure that the result is usable and safe, and reliably minimizes negative and unintended consequences? By considering the multiple use cases and potential unintended consequences of AI, Michigan Tech can advance its application throughout the product lifecycle, from initial concept through development and validation, to end-of-life management.
What will be the role of humans in software engineering when AI can write code faster and more correctly than humans?
Parents of prospective students frequently ask us whether programming will continue to be a viable career option in the age of AI. It’s a reasonable question, because AI is already democratizing software development. In the past, people learned to program by starting with the simplest 3-4 line “Hello World” programs and then adding complexity from there. With AI, one can easily upgrade to more complex and engaging software from the start. In a world where teenagers can create engaging and fun video games after a brief introduction to coding and lots of trial and error with ChatGPT, it’s easy to see why parents are concerned about the viability of programming and game development. software as a career.
Our response to concerned parents: fear not, designers and developers will always have a role to play. However, as software platforms become more complex, the developer’s job will be profoundly different.
AI will increasingly be used to create code, particularly simple routines and APIs (interfaces between distinct parts of the application). Humans will be needed to strategize how to scale and maintain complex codes enterprise-wide. Dianne Marsh ’86 ’92 (BSMS Computer Science) made this point in a talk she gave at Michigan Tech a few years ago. She pointed out that the code at Netflix, where she was then director of consumer product safety and trust, has become so complex that no one really understands everything it does. As such, developers are increasingly focusing on creating and implementing test cases to assess code quality and its impact on the overall system. The proliferation of AI-generated code will only accelerate this trend across all industries.
How can we ensure that AI is ethical and does not perpetuate existing inequalities?
Amazon recently backed away from using natural language processing to review job applications because it favored words more commonly used by men. This is just one of many troubling examples of AI bias causing problems in society. Equally disconcerting is the tendency of large language models (LLMs) like ChatGPT to hallucinate, stating as fact things that simply aren’t true. Currently, when LLMs are wrong, they are wrong with certainty, meaning there is no warning to the user that one claim is less reliable than another. These problems are already limiting the usefulness of AI and eroding trust in it. Even the recent winners of the 2024 Nobel Prize in Physics – recognized for their seminal work leading to the development of artificial intelligence – used the announcement as an opportunity to warn on the potentially negative impacts of AI.
The scope of AI is already vast; it will be ubiquitous by 2035. However, once the novelty of AI wears off and the hype dies down, its presence will likely be less obvious to us, as it will take a back seat as a technology enabling. As this happens, we need to focus even more on assessing AI bias and its unintended impacts. Faced with the science fiction limits of AI “machine takeover,” we must also ensure that AI respects human life above all. Just as we train software developers to integrate security into their design, and just as we train mechanical engineers to think about manufacturability and maintainability, we must also train AI developers to take into account the ethical and societal impacts of their AI creations.
*****
While AI presents many challenges, it also presents wonderful opportunities. AI has already made writing more efficient, streamlined supply chains, and demonstrated remarkable creativity. AI also does a very good job of predicting and identifying, skills it will only improve over time. For example, AI-based image analysis tools promise to identify tumors where oncologists cannot. AI can also reliably predict a protein’s three-dimensional structure from its one-dimensional sequence of amino acid building blocks, solving a decades-old problem that many biophysicists thought could never be solved.
Nevertheless, our concern and caution should match our enthusiasm. Developing methods to identify and address these issues at early stages of development must be a cornerstone of our academic programs. These complex issues require diverse experiences and expertise, meaning design will no longer be solely the domain of technical professionals. Humanists, social scientists and others must be involved from concept to realization. The need to prepare our engineers and computer scientists to collaborate effectively in diverse teams will make experiential opportunities like Michigan Tech’s enterprise program even more important in the future.
Michigan Tech has historically done a good job addressing the societal impacts of technology, and we will draw on this expertise to address the challenges posed by AI. We will continue to offer and expand experiential learning opportunities where computing and AI take center stage. Using the agility that Michigan Tech is known for, we will continue to evolve and adapt to unknown unknowns.
Michigan Technological University is a public research university founded in 1885 in Houghton, Michigan, and is home to more than 7,000 students from 55 countries around the world. Consistently ranked among the nation’s top universities in terms of return on investment, Michigan’s flagship technological university offers more than 120 undergraduate And diploma degree programs in science and technology, engineering, computer science, forestry, business and economics, health professions, humanities, mathematics, social sciences and the arts. The rural campus is located just a few miles from Lake Superior in Michigan’s Upper Peninsula, offering year-round opportunities for outdoor adventure.
