An Interview With Dr. Craig Froehle


AI is advancing rapidly, raising questions about how it will transform the workplace.

Large-language models like GPT-4 can generate articles, emails, and code, making many wonder if specific jobs could soon become automated.

To dig deeper into the potential impact of AI on jobs, we spoke with Craig Froehle, Ph.D., a professor at the University of Cincinnati’s Lindner College of Business.

Froehle, who shared valuable insights on AI’s role in the future of work in a recent University of Cincinnati article, brings a nuanced perspective on this critical issue.

In our exclusive interview, Froehle addresses the following subjects:

  • AI’s current and future role in white-collar jobs.
  • Jobs most vulnerable to AI.
  • A need for lifelong learning in an AI-driven world.
  • Ethical considerations surrounding AI.
  • Concerns about bias and the importance of transparency.

This article delves into the insights from Froehle’s interview, exploring the impact of AI on digital marketing and SEO professions, the emerging roles in the AI economy, and the responsibilities of both workers and businesses in this new era of work.

AI Won’t Eliminate Jobs But Will Reshape Many

While AI tools can mimic human writing and creativity, Froehle believes they will only partially replace most jobs after a while.

“These models will likely make white-collar workers more productive because they generate content quickly, but aren’t eliminating many jobs anytime soon. Tasks within jobs? Yes, some. But whole jobs? Less likely.”

According to Froehle, AI is best at repetitive, rules-based tasks that don’t require original thinking.

For now, creative work and critical analysis still demand human oversight.

“Somebody with training and experience needs to look at what the model produced and evaluate its fitness for use.”

That means AI will reshape jobs more than eradicate them. The tools may handle rote tasks, freeing workers’ time for higher-value responsibilities.

AI Can’t Replace Thinking

Despite rapid advancements in AI tools like OpenAI’s GPT-4 and Google’s Bard, Froehle emphasized that these AI models are not honestly “thinking.”

“These tools are already beneficial adjuncts to many tasks that fill up white-collar workers’ days. But creative tasks and those that require critical thinking, analysis, and even a lot of technical work still require someone with expertise to verify the quality of the model’s generated output.”

In other words, AI isn’t set to replace jobs entirely, but to augment specific tasks within jobs, particularly those that are more procedural and less creative.

The AI Job Risk Spectrum

When asked which jobs are most vulnerable, Froehle pointed to coding and content creation.

“Any task where one could generate the content primarily, if not solely, by copying or emulating what’s already readily available on the internet, is easy pickings for these LLMs. Basic coding is an excellent example of this.”

The same goes for basic text and image generation, like social media posts, as the rules are relatively formulaic.

“Tasks that are low on originality – that produce content according to fairly replicable rules or guidance – are probably the ones that will be handed off to AI tools most rapidly.”

It’s important to note, however, that AI still hasn’t perfected the replication of a brand’s (or individual’s) or unique voice.

Responsible AI Deployment

Froehle urges diligence around training data and output validation as companies race to implement AI.

He warns that scraping the web for training data could violate copyrights and intellectual property rights.

“If [comedian Sarah Silverman] is successful, it will open a significant can of worms for AI model builders,” said Froehle, referencing Silverman’s recent lawsuit against OpenAI over GPT-3’s use of her material.

Froehle also cautioned that users must carefully review AI-generated content before releasing it publicly.

Thorough validation processes are essential to prevent reputational and legal risks.

“Depending on the size and complexity of the content, that task could be challenging indeed, and we don’t currently have good tools to automate that inspection.”

One solution, Froehle points out, is to build a model using only your company’s internal data.

“A recent trend has been companies using their internal data for training models that they, and only they, will use, which could be much more common moving forward.”

The Future-Proof Worker

To stay in demand amidst AI adoption, Froehle urges constant skill-building and adaptation.

“Someone who wants to ‘future-proof’ their role in a company better be as adaptive as possible.”

He believes companies must also invest heavily in retraining employees for new responsibilities.

“Organizations committed to their employees will need to invest in employee training and up-skilling, and they’ll need to work harder than ever to help move employees into roles that match their capabilities.”

Partnerships with colleges could provide the necessary training programs.

Froehle warns of the risks of investing in something other than appropriate training.

“As fast as AI tools are likely to evolve, I can imagine companies who don’t invest that way start to see their employee turnover expenses double, or more, in a very short period.”

The Emerging AI Economy

Froehle pointed out the emergence of new jobs such as prompt engineering, AI model builders, and content validators as AI becomes more advanced and commonplace.

He also highlighted the potential benefits of AI, such as increased efficiency for workers if businesses can rethink the role of people in their processes.

However, he expressed concern over the potential misuse of AI.

“We know that false information is spread more quickly than truthful information, at least via social media, so bad actors could easily start destabilizing modern society by capitalizing on the general public’s lack of ability … or interest in critical assessment of what they see, hear, and read before passing it along to their family, friends, and followers.”

Despite his concerns, Froehle remains optimistic about the future of work with AI.

“I think the upside potential of these models, and the next umpteen generations we’ll see in the coming years, is hugely compelling. It would be great if informed legislation could be devised to amplify the pros of generative AI while tamping down its cons as much as we can manage.”

As AI continues to evolve, the future of work will undoubtedly transform. The key lies in adapting and leveraging these tools to enhance our capabilities rather than replace them entirely.

The Verdict

If appropriately leveraged, AI can augment human capabilities. However, it still doesn’t fully replicate creativity and critical thinking.

With thoughtful implementation, AI can increase efficiency without rendering workers obsolete.

The verdict is that AI will not replace jobs, but reshape many, by automating repetitive tasks.

Workers must adapt through constant learning, and companies should invest in retraining to avoid turnover.

New roles like prompt engineers are emerging, but ethical risks around bias and misuse necessitate responsible AI adoption.


Featured Image: Stokkete/Shutterstock





منبع