3 Things About AI and the Future of Work – Inside Higher Ed

This post was originally published on this site.

Since the public release of ChatGPT in late 2022, artificial intelligence has rocketed from relative obscurity to near ubiquity. The rate of adoption for generative AI tools has outpaced that of personal computers and the internet. There is widespread optimism that, on one hand, AI will generate economic growth, spur innovation and elevate the role of quintessential “human work.” On the other hand, there’s palpable anxiety that AI will disrupt the economy through workforce automation and exacerbate pre-existing inequities.

History shows that education and training are key factors for weathering economic volatility. Yet, it is not entirely clear how postsecondary education providers can equip learners with the resources they need to thrive in an increasingly AI-driven workforce.

Here at the University of Tennessee, Knoxville’s Education Research and Opportunity Center, we are leading a three-year study in partnership with the Tennessee Board of Regents, Advance CTE and the Association for Career and Technical Education to explore this very subject. So far, we have interviewed more than 20 experts in AI, labor economics, career and technical education (CTE), and workforce development. Here are three things you should know.

1. Generative AI is the present, not the future.

First, AI is not new. ChatGPT continues to captivate attention because of its striking ability to reason, write and speak like a human. Yet, the science of developing machines and systems to mimic human functions has existed for decades. Many people are hearing about machine learning for the first time, but it has powered their Netflix recommendations for years. That said, generative AI does represent a leap forward—a big one. Simple machine learning cannot compose a concerto, write and debug computer code, or generate a grocery list for your family. Generative AI can do all of these things and infinitely more. It certainly feels futuristic, but it is not; AI is the present. And the generative AI of the present is not the AI of tomorrow.

Our interviews with experts have made clear that no one knows where AI will be in 15, 10 or even five years, but the consensus predicts the pace of change will be dramatic. How can students, education providers and employers keep up?

First, we cannot get hung up on specific tools, applications or use cases. The solution is not simply to incorporate ChatGPT in the classroom, though this is a fine starting point. We are in a speeding vehicle; our focus out the window needs to be on the surrounding landscape, not the passing objects. We need education policies that promote organizational efficiency, incentivize innovation and strengthen public-private partnerships. We need educational leadership focused on the processes, infrastructure and resources required to rapidly deploy technologies, break down disciplinary silos and guarantee learner safeguards. We need systemic and sustained professional development and training for incumbent faculty, and we need to reimagine how we prepare and hire new faculty. In short, we need to focus on building more agile, more adaptable, less siloed and less reactive institutions and classrooms because generative AI as we know it is not the future; AI is a harbinger of what is to come.

2. Focus on skills, not jobs.

It is exceedingly difficult to predict which individual occupations will be impacted—positively or negatively—by AI. We simply cannot know for certain whether surgeons or meat slaughterers are at greatest risk of AI-driven automation. Not only is it guesswork, but it is also flawed thinking, rooted in a misunderstanding of how technology impacts work. Tasks constitute jobs, jobs constitute occupations and occupations constitute industries. Lessons from prior technological innovations tell us that technologies act on tasks directly, and occupations only indirectly. If, for example, the human skill required to complete a number of job-related tasks can be substituted by smart machines, the skill composition of the occupation will change. An entire occupation can be eliminated if a sufficiently high share of the skills can be automated by machines. That said, it is equally true (and likely) that new technologies can shift the skill composition of an occupation in a way that actually enhances the demand for human workers. Shifts in demands for skills within the labor market can even generate entirely new jobs. The point is that the traditional approach to thinking of education in terms of majors, courses and degrees does learners a disservice.

By contrast, our focus needs to be on the skills learners acquire, regardless of discipline or degree pathway. A predictable response to the rise of AI is to funnel more learners into STEM and other supposed AI-ready majors. But our conversations, along with existing research, suggest learners can benefit equally from majoring in liberal studies or art history so long as they are equipped with in-demand skills that cannot (yet) be substituted by smart machines.

We can no longer allow disciplines to “own” certain skills. Every student, across every area of study, must be equipped with both technical and transferable skills. Technical skills allow learners to perform occupation-specific tasks. Transferable skills—such as critical thinking, adaptability and creativity—transcend occupations and technologies and position learners for the “work of the future.” To nurture this transition, we need innovative approaches to packaging and delivering education and training. Institutional leaders can help by equipping faculty with professional development resources and incentives to break out of disciplinary silos. We also need to reconsider current approaches to institutional- and course-level assessment. Accreditors can help by pushing institutions to think beyond traditional metrics of institutional effectiveness.

3. AI itself is a skill, and one you need to have.

From our conversations with experts, one realization is apparent: There are few corners of the workforce that will be left untouched by AI. Sure, AI is not (yet) able to unclog a drain, take wedding photos, install or repair jet engines, trim trees, or create a nurturing kindergarten classroom environment. But AI will, if it has not already, change the ways in which these jobs are performed. For example, AI-powered software can analyze plumbing system data to predict problems, such as water leaks, before they happen. AI tools can similarly analyze aircraft systems, sensors and maintenance records to predict aircraft maintenance needs before they become hazardous, minimizing aircraft downtime. There is a viable AI use case for every industry now. The key factor for thriving in the AI economy is, therefore, the ability to use AI effectively and critically regardless of one’s occupation or industry.

AI is good, but it is not yet perfect. Jobs still require human oversight. Discerning the quality of sources or synthesizing contradictory viewpoints to make meaningful judgments remain uniquely human skills that cut across all occupations and industries. To thrive in the present and future of work, we must embrace and nurture this skill set while effectively collaborating with AI technology. This effective collaboration itself is a skill.

To usher in this paradigm shift, we need federal- and state-level policymakers to prioritize AI user privacy and safety so tools can be trusted and deployed rapidly to classrooms across the country. It is also imperative that we make a generational investment in applied research in human-AI interaction so we can identify and scale best practices. In the classroom, students need comprehensive exposure to and experience with AI at the beginnings and ends of their programs. It is a valuable skill to work well with others, and in a modern era, it is equally necessary to work well with machines. Paraphrasing Jensen Huang, the CEO of Nvidia: Students are not going to lose their jobs to AI; they will lose their jobs to someone who uses AI.