Despite the excitement about emerging consumer-grade AI like ChatGPT, Bard, and Bing, students are rather gloomy about what these advancements mean for their future. During a recent survey of students aged 16-61, nearly one-third expressed worry about the impact of AI on their career or potential career. That’s an understandable sentiment, given that large language model (LLM)-based AI platforms can already Google faster than any human and autonomously perform administrative tasks like managing calendars and responding to emails.
On the flip side, students have a clear-eyed assessment of what AI may or may not be capable of doing. About four in ten respondents think AI can “generate accurate and reliable results,” and almost two-thirds say, “AI can’t replace human intelligence or creativity.” These responses provide a healthy dose of realism about AI’s prospects and how it might impact those seeking to make their own mark in the professional world.
Students’ nuanced view of AI demonstrates what many of us already know. The ultimate goal of AI may be unsupervised production. However, some of the most exciting AI today still needs human supervision. Despite concerns about AI replacing roles, the complete takeover of jobs is not entirely unavoidable. That said, there are tasks for which AI is well suited, and as algorithms get smarter and developers fine-tune their models, AI will make its way into more functions in the near, medium, and far terms.
We sat down with Mike Krause, Senior Data Science Director; and Ari Kamlani, Principal Data Scientist, to learn more about their AI predictions in the workplace.
The story of the next one to five years is one of AI helping humans become more productive. Both Mike and Ari predict that more administrative tasks, from data management to paralegal activities, will become the domain of AI-enabled tools.
The other major arena for AI will be in automation.
“In the next 1-3 years,” Mike predicts, “AI will replace factory and warehouse work driven by increased use of robotics and AI to automate processes. Various roles like quality assurance in factories and software engineering will also be replaced when LLMs can be used to effectively test code.”
According to Ari, standard business processes or workflows that can be codified and have low variability and high consistency can be replaced by AI. In many cases, AI is already very present in these areas.
“We tend to associate these business processes with technologies such as RPA, either powered by traditional backend rules-engines or now modern AI,” adds Ari.
As we move a bit further into the future, Ari and Mike see AIs blurring the lines between human and machine content output and further evolving in places they already exist. “AI tools will continue to be integrated into the devices we use every day; people will be using AI without even realizing it (like your e-mail autofill),” Mike predicts.
The TL;DR of near-term AI in the job market is that it will find a niche automating many administrative tasks—a role that will allow fewer humans to do more.
Within five years, both Ari and Mike envision AI moving towards the edge as it evolves. Mike believes AI will be embedded in many more devices such as traffic cameras, pollution sensors, and vehicle safety equipment. Additionally, Ari sees AI addressing the limitations of resource-constrained devices, specifically the performance gap between the cloud and various classes of connected edge device footprints [being] minimized.
As AI becomes more mature and reliable, it will become more adept at handling various activities at the network’s edge. Many of these roles will help keep humans safer, whether by keeping humans away from dangerous jobs in warehouses or improving things such as security and autonomous driving.
The medium-term outlook for AI in the jobs market is one of learning and maturing. Ari notes that we deploy our models without knowing if they will have a true effect in the real world. In some application contexts, there may be long lag periods before we are able to realize their true value. In the next decade, the gap between “real-world environment utility and the proxies used for population sampling and AI model training” will shrink as AI improves how it helps humans in very specific roles.
In the long term, AI will likely evolve from a specialist to a generalist, as it finds ways to augment and enhance the human experience holistically.
Mike predicts that AI will be able to integrate numerous data streams to recommend or implement end-to-end strategic operations for applications such as scheduling and planning military operations, corporate planning, government policy, supply chain management, and investment strategy.
Ari believes that rather than having very disparate entities people interact with, for example, an app for this, an app for that, they will be stitched together as cohesive integrated experiences.
This evolution will allow AI to aid humans in more complex tasks, from surgical procedures to judicial proceedings. Whatever the role, AI in the more-distant future promises to fundamentally change the way we interact with technology, whether, as Ari notes, “via software agents or real physical robots as companions to our lifestyle.”
Should these predictions prove prescient, the presence of AI in the job market will be complicated. It will soon become an excellent assistant, presiding over administrative tasks that many humans likely find tedious or cumbersome. Even for slightly more advanced tasks, like copywriting or e-discovery, humans may find AI to simply be more efficient. Yet, there is no reason for people in those types of roles to fret. AI is likely to provide plenty of opportunities as it helps humans multiply their efficiency.
Ari likens the next tranche of technology-driven opportunity to the current creator or gig economies. “With the explosion of generative AI, particularly with the inclusion of mixed modalities (e.g., text-to-image, image-to-text), these ‘creator’ economies shall be on the rise. This could mean specialization opportunities for new startups and/or providing alternative possibilities to larger firm offerings.”
Mike sees supervisory roles for humans as AI pervades over replacing more roles. Especially when these roles are coupled with robotic operations, there will be a need for human oversight. For instance, “content moderators” will be needed to control and flag autonomous AI systems.” There will also be a need for “prompt engineers” for jobs like interacting with autonomous drones and vehicles, content creation, engineering fields, and cleaning robots.
Most (61%) current students say AI tools like ChatGPT will become the new normal. And, while this new normal may be rife with AI performing more mundane tasks for humans, it will also be blossoming with fields of potential. Students or those considering career changes may want to focus on AI oversight and enhancing their own productivity with growing models. The more we understand how AI works and how to work with it, the more likely it is that AI continues to be a strong complement to our career objectives.
LEARN MORE ABOUT BEYOND LIMITS COGNITIVE AI AND EXPLAINABLE AI SOLUTIONS.
This website uses cookies. By continuing to use this website or by clicking “Accept All Cookies,” you are giving consent to cookies being used. For more information on cookies and how you can disable them visit our Cookie Policy.