You have /5 articles left.
Sign up for a free account or log in.
I just returned from the largest-ever UPCEA annual conference in Washington, D.C. More than 1,000 leaders in professional, continuing and online education gathered to learn, share and prepare for the future. I had the privilege to greet hundreds of those as they registered for the conference on Wednesday. Many of those attending were excited about the potential of new knowledge that will be provided by the emerging generative technologies. Yet, others were somewhat fearful of the direct impact of generative AI on current and future jobs for themselves and their colleagues. Everyone seemed interested in just how ChatGPT works and how it is likely to change over time. Let’s sort through the current issues and concerns, then suggest what we are most likely to see in the near future.
It is certainly true that GPT-4 and other large language models bring new abilities and opportunities to higher education. To quote Udacity CEO and co-founder, Stanford professor Sebastian Thrun, generative AI is “The single biggest thing we are ever going to see in our lifetime.” The reach of this technology goes much farther than education. Our society will soon become populated with personal assistants serving individual and group needs across our culture. They will become high-powered personal assistants, tracking our individual interests, preparing us for work daily and supporting us in advancing our careers. In effect, each and every one of us will have a supercharged, highly empowered “Siri” or “hey Google” that will be proactive and perform much more than simple reminders, fact checks and search engine lookups. The next generation of personal assistants will anticipate our knowledge needs and provide value adds in information discovered through deep searches every day.
As GPT-4 was released, OpenAI also released a treasure trove of benchmarking and further testing of their most recent versions. Incredibly, the GPT-4 version stacks up very well on standardized tests against some of our top graduate assistants. Check out these scores: Uniform Bar Exam 90th percentile; SAT evidence-based reading/writing 93rd percentile; SAT Math 89th percentile; GRE Verbal 99th percentile, with dozens more documented and presented. To me, these and the rest of human-AI benchmarking results that OpenAI reported are astounding. I encourage all readers to check out the report that can be found online.
As mentioned in “Online: Trending Now” earlier this month, Ed Felten (Princeton), Manav Raj (University of Pennsylvania) and Robert Seamans (New York University) have released a soon-to-be peer-reviewed paper—“How Will Language Modelers Such as ChatGPT Affect Occupations and Industries?”—in which the case is made that generative AI has the potential to impact teaching in many fields.
Importantly, we in higher education will determine if these tools will supplement or, rather, replace workers in our field. If we follow the early lead of businesses, it is likely that we will reduce our staffing while expanding the use of generative AI. The question has been posed in substantial early surveys. Resume-Builder reports:
In February, ResumeBuilder.com surveyed 1,000 U.S. business leaders to see how many companies currently use or plan to use ChatGPT … 1 in 2 companies have replaced workers with ChatGPT since November. Forty-eight percent of companies have replaced workers with ChatGPT since it became available in November of last year. In the future, ChatGPT may lead to more layoffs according to business leaders. When asked if ChatGPT will lead to any workers being laid off by the end of 2023, 33% of business leaders say ‘definitely,’ while 26% say ‘probably.’ Within 5 years, 63% of business leaders say ChatGPT will ‘definitely’ (32%) or ‘probably’ (31%) lead to workers being laid off.
In addition, respondents in the Resume-Builder survey indicated that nine in 10 business leaders say it’s beneficial for job applicants to have ChatGPT experience and that it’s possible having ChatGPT experience could help workers keep their jobs. In assessing candidates to hire, 92 percent of business leaders say AI/chat bot experience is a plus.
The advice seems clear. We all may be well served by investing some time in understanding this technology and testing out ways in which we may be able to use it to advance our work.
The question of just how this large language model works is perhaps best described by ChatGPT itself. I asked it to describe the process. In a response to my question on March 17, 2023, ChatGPT wrote, “Chatbot GPT (ChatGPT) is a language model based on the GPT-4 architecture developed by OpenAI. It functions by utilizing deep learning techniques, specifically, a type of artificial neural network called a Transformer. The GPT in ChatGPT stands for “Generative Pre-trained Transformer.” The app went on to describe six key steps in a “high-level overview.” Those include pre-training, fine-tuning, tokenization, context vector, attention mechanism and decoding. The complete reply can be found in my presentation notes as well as—if you prefer—a human-written summary from March 10, 2023, “How does ChatGPT work?” by David Gewirtz, ZDNet.
Sharon Goldman writing in VentureBeat highlights the “hottest” new trend in generative AI. It is the rapidly growing number of productivity apps. We are likely to see this trend continue to play out through the rest of the year. These tools have the possibility of saving time while providing higher-quality results in many different aspects of your work life. Many are free or low-cost.
Looking ahead to the not-too-distant future, we are likely to see more about biocomputers. Using brain and stem cells, researchers are hoping to vastly accelerate and expand performance. Research is already well under way at Johns Hopkins:
It’s called organoid intelligence, or OI and it uses actual human brain cells to make computing “more brain-like.” OI revolves around using organoids, or clusters of living tissue grown from stem cells that behave similarly to organs, as biological hardware that powers algorithmic systems. The hope—over at Johns Hopkins, at least—is that it’ll facilitate more advanced learning than a conventional computer can, resulting in richer feedback and better decision-making than AI can provide … Using human brain cells to power computers has obvious ethical implications, which the researchers openly acknowledge.
Decades ago, a wave of AI replaced assembly-line blue-collar jobs. This version of AI is more likely to replace administrative white-collar jobs that previously required college, even graduate degree level knowledge and skills. However, just as was the case those decades ago, the overall number of jobs increased; that is likely to happen again with the advent of generative AI. The new jobs will likely require a thorough understanding of generative AI and the skills to apply this emerging technology to advance effectiveness and efficiency.
