When ChatGPT entered the world, headlines about it felt impossible to escape. Fears soon emerged that this robot could eliminate most or all jobs – and artificial intelligence (AI) has only become an even hotter topic since then. There’s a twist, though: According to recent research, although many people see AI as a threat, they nevertheless use it a ton at work.
You might wonder: How can we simultaneously fear and embrace AI in the workplace? What does this say about the future of work and the role AI will play? We’ve explored this question in depth below.
There’s no doubt artificial intelligence is transforming businesses. But can it take employees’ jobs?
Among the 1,000 employees Resume Genius surveyed for a 2023 study, 69 percent worry that AI could take over their jobs entirely. Similarly, 74 percent predicted that AI will eliminate all forms of human labor, and 37 percent expected a reduction in job opportunities as AI grows. Additionally, 63 percent of respondents said they fear AI, and 58 percent said they find AI dangerous.
However, not all employees feel that AI is coming for their jobs. In fact, 25 percent of Resume Genius respondents would rather interact with an AI tool at work than an actual person. Additionally, 38 percent of respondents said AI would create more, not fewer, job opportunities. Another 25 percent said AI would have no impact on job availability.
These responses suggest that not all workers fear robots taking their jobs; in fact, many employees see this expanding technology as a boon.
Resume Genius found that AI use at work is already widespread. In fact, 75 percent of respondents said they feel positive about using AI in work settings. Expanding on that, 83 percent of respondents said they trust AI, and 82 percent said they find AI helpful. That same percentage (82 percent) reported currently using ChatGPT or other AI chatbots and tools at work.
Interestingly, an even higher percentage (94 percent) said they use AI to complete specific work tasks. On top of that, 85 percent use AI in place of Google and other search engines. Among these AI users, 87 percent have admitted their AI usage to co-workers and supervisors. It’s clear that AI isn’t just the future of work – it’s a small business tech trend that’s making its mark right here, right now.
“The future of AI is in what we call ‘cobots,’ or collaborative robots,” explained Shawn David, creator of the course Automate to Win. “These are simple algorithms that will help you do things like remember a recipe for the cake you are making in real time, up to and including a J.A.R.V.I.S.-like assistant that is fully capable of anything short of quantum computing.” Based on the Resume Genius findings, employees may be in the early stages of implementing cobots in the workplace.
Resume Genius respondents named over a dozen jobs they see as “safe” from a potential AI takeover, including chefs, artists and writers. Many of these jobs have one thing in common that safeguards them from a potential AI takeover.
“If you are a product builder, your job is very safe,” explained Tim Mudd, webflow developer at and founder of Mudd Media. “That product could be a work of journalism, a website, an app, [or] a legal brief. AI is not going to replace you anytime soon.”
Foo Conner, founder-in-residence at Carnegie Mellon University’s Swartz Center for Entrepreneurship, agreed. “The jobs at risk are boring and repetitive,” Conner pointed out. “[For example,] for companies with large pools of images, AI can go through to catalog, organize [the images] and keyword them. One company is helping newsrooms preserve their archives and modernize them by tagging articles and images with the use of AI.”
Resume Genius asked survey respondents which jobs they expected AI to take over entirely. They gave over a dozen answers, including receptionists, telemarketers and cashiers.
According to Mudd, “competing against the output of a prompt” puts you at risk.
Resume Genius also asked respondents to name the three industries they expected AI to take over entirely. The leading answer was information technology (IT) at 46 percent, followed by manufacturing at 31 percent and healthcare at 27 percent.
However, Derrick Mains, CEO and founder of The Process Fixer, disagreed on the likelihood of AI overtaking the manufacturing industry. He thinks blue-collar, manual labor work is safe – for now – from an AI takeover. “Robotics costs are too high and the ROI too low for companies to consider automating mundane tasks,” Mains explained. “This is true of most industries, except those with quality concerns, such as aerospace, engineering and healthcare. Those industries have already been leading with automation to remove mistakes and improve quality as well as reduce costs.”
Based on our conversations with AI experts, here are some do’s and don’ts for using AI at work.
Although much fuss has been made over ChatGPT writing full-length web articles, the quality of these articles can be far below what a human can deliver. However, ChatGPT is sharp enough to create somewhat meaningful outlines or other documents necessary for plotting a project.
“[AI] is very helpful in providing outlines and rough drafts, and in situations where coming off as conventional is desirable,” Mudd noted. This could include creating simple, rigidly formatted documents that are repetitive and painstaking to assemble.
Even mathematicians make numerical errors. That’s the thing with people: Even the best of us inevitably make mistakes. AI, though, is faster and more accurate when it comes to computation.
“Any process that has manual tasks that can be repeated and have a high risk of error are the ones that AI can step in and take over,” David explained. He added that AI doesn’t “have transposition errors, so things like bank wire transfers with no takebacks are prime for automation and AI usage.”
According to David, data-based processes like data segmentation, qualification and quantification as well as data manipulation, categorization and curation are ripe for AI intervention.
“For the most part,” Mains noted, “we are seeing AIs applied to numerical data doing really good work and improving over time. Accounting, business intelligence, reporting, supply levels and inventory are all areas that seem to be improving rapidly.”
Reading the fine print on contracts, policies and agreements is tedious but necessary. Businesses are increasingly using AI tools to review these documents and serve as an extra set of “eyes” to spot crucial caveats, loopholes and mistakes.
“Several of the startups [I work] with … are using custom language models to review contracts [and] insurance policies, and find things that might otherwise be overlooked,” Conner said. “Law firms don’t have unlimited budgets for text reviews, [and] an extra set of ‘eyes’ can help.”
AI is not quite as sharp as a human writer or editor. It may plagiarize or pass off inaccuracies as fact. You may consider using AI as a writing assistant, but you should be wary about letting it do all the work for you. And you should always edit anything AI generates very closely.
“If you ask ChatGPT to write you a blog post, screenplay or other creative work, it will likely write you a pretty bad one,” Mudd said. “It is competent but ultimately generic and derivative in its output.”
Mains agrees. “Although there is a lot of talk about AI in writing, my own AI work in writing shows that, if given the opportunity and without regular human input on direction, AI will go rogue and continue down paths that get more and more divergent from the original intent,” Mains noted.
Using tech to gain a competitive advantage is one thing, but be sure you’re using it correctly. For example, tech publication CNET was roundly blasted when it used AI to write dozens of error-riddled articles.
Just as AI isn’t quite ready to write anything truly meaningful, it isn’t yet sharp enough to devise your business strategy. Sure, you can use AI to gather key business data and perhaps, if you’re starting a new business, outline your business plan before you write it. But it’s not nearly ready to decide how you should go about marketing and selling your products or services.
“Further-level strategy is not there yet, but the base level of segments is there,” David explained. In other words, AI can help you get the numbers and facts you need, but it can’t yet decide what should be done with them. At least for now, that’s still where you come in.
AI can’t reach out to people in your network. In its current state, it can interact only with its user (you). It also can’t read your mind and absorb your expertise into whatever algorithms it uses. This is why using AI to assist – not take over – your work tasks is essential.
“AI is no replacement for expertise and credentials,” Conner cautioned. “We’ve read the stories of lawyers taking shortcuts and being reprimanded by judges for using ChatGPT. [AI] cannot argue in court or give you the network of peers to call for the hard problems.”
In specific industries, AI poses enough of a potential threat that some businesses require their employees to sign agreements forbidding its usage. For example, as part of their employee handbooks, content writing agencies may ban writers from using ChatGPT. This trend raises the question: Should your company create a workplace ethics policy surrounding AI use? According to Conner, the answer is yes.
“As AI evolves, companies should establish ethics policies that guide employees’ use of AI,” Conner advised. “AI is not a catch-all. Companies that over-rely on AI will lose clients. Humans will always be required for the process.”
However, Mains recommends creating an AI document somewhat different from a company policy.
“Map out all of your processes,” Mains suggested. “Not just sales and marketing. Map every single button click and task for all areas of the organization.” Then, he said, “Create a systems map for your company and determine all the software you use. … Do a time study on all those process maps and determine how much time each takes. Analyze the data on what takes the most time and takes the least effort to correct.”
Mains also suggests creating use cases. “Create use cases on how those inefficient areas might be able to be automated by existing tech and AI,” Mains said. “Do small beta tests with AI tools to see if they improve the workflow.” From there, you’ll have a written document – something at least adjacent to a policy – detailing where AI use within your company makes sense.
For now, the average employee doesn’t have much to worry about when it comes to AI, and certain professions might entirely resist AI takeovers. Plus, in its current state, AI can save your employees time without reducing their work quality or entirely replacing your team of actual people. However, Mains noted that what’s true about AI today could change sooner rather than later.
“I currently do not suggest that any of my clients incorporate AI as a fundamental system in their organization,” Mains said. The key here is “currently.” According to Mains, at the moment, “What we are seeing in the marketplace is not a good representation of AI. Right now is the time to get prepared.” The true race may be in the near future.