In its new A.I. Future of Work report, LinkedIn took a look at how different sectors are likely to adapt to the new technology. It analyzed some of the most common occupations on its platform, predicting which of them would have a stronger likelihood of changing by identifying which skills could be augmented by A.I. and which skills could be done by humans. 

Software engineer stands atop the list with the “greatest number of skills potentially augmentable by GAI,” at 96%. That’s all to say, like a tech spin on Annie Get Your Gun, anything a software engineer can do a robot can do better (or at least can do most of—as in the song, it’s debatable). Forward-facing employees like customer service representatives, salespeople, and cashiers are next on the list, with LinkedIn predicting that generative tools could replicate skills like customer support, cold calling, merchandising, and typing. Still, these roles involve human-based tasks like stocking, waiting tables, or food and hospitality service, all more human-based skills—so don’t expect a Roomba waiter yet. 

This could all sound like proof for those who are worried that ChatGPT will replace their jobs. After all, much like the story of Kronos, the field that pioneered A.I. invention could be hit by its own innovation—a report from Challenger, Gray & Christmas found that 3,900 employees were laid off due to artificial intelligence, most of the affected were in the tech sector where software engineers work. But it’s a tale of evolution rather than extinction, as LinkedIn suggests that generative A.I. could help workers be more productive as it takes over less human-based skills. 

“The skills necessary to do one’s jobs are always evolving with the degree of change varying from job to job and industry to industry. What our data suggest at LinkedIn is that A.I. won’t replace our jobs, but it will accelerate the change in skills required to do those jobs,” Karin Kimbrough, LinkedIn’s chief economist, says in a statement sent to Fortune. This might look like someone in sales changing their priorities from “cold calling to deepening sales relationships,” Kimbrough adds, explaining that “ultimately the skills required for those jobs augmented by A.I. will change, rather than completely going away, and people skills will become even more valuable.”

For example, the report explains, if A.I. can begin writing code in different programming languages for software engineers, then they can focus working with project management tools like Jira to improve their communication with key stakeholders. Likewise, teachers could become more productive if A.I. helps them with lesson planning and curriculum development, enabling them to give more individual attention to students and focus on classroom management.

Some experts say that rather than killing off jobs, A.I. is more likely to change the workflow. “A.I., as far as I see it, doesn’t necessarily replace humans, but rather enhances the work of humans,” Dan Wang, a Columbia Business School professor, told CNN. 

A report by ServiceNow and Pearson finds that both doomsday (i.e., job extinction) and optimistic scenarios (i.e., job evolution) are plausible, depending on where you live and what industry you work in. Since generative A.I. is still so new, the  workforce is justifiably ambivalent about A.I.; a recent Ipsos Global Advisor survey found that across 31 countries, 52% of adults are nervous about A.I. services and 54% are excited about them. A fair number (57%) think A.I. will change their job, and a smaller but still significant group (36%) expect their role to be replaced by it.

While the iron regarding ChatGPT and other services was running hot at the start of 2023, reports of errors or perhaps simply dwindling interest has led to a dip in traffic in chatbot usage for the first time this summer. Still, consulting groups like McKinsey predict that A.I. will accelerate a pandemic-era trend of workers changing sectors, with people in consumer-facing roles more likely to change industries than those in health care or jobs with skills less able to be taken over by A.I. Or, as LinkedIn’s analysis shows, they’ll just shift their day-to-day activities.

These are the top five jobs LinkedIn predicts will be affected by A.I., ranked by the percentage of skills most likely to be potentially augmented by the tools: 

Software engineer: 96% 
Customer service rep: 76%
Salesperson: 59%
Cashier: 59%
Teacher: 45%

And these are the jobs with the fewest number of skills able to be mimicked or strengthened by A.I.: 

Oil field operator: 1%
Environmental health safety specialist: 3%
Nurse: 6%
Medical doctor: 7%
Driver: 9%