Cynic Tool

[Denisha Simcoe]

Artificial intelligence tools such as ChatGPT have drawn the attention of University faculty, according to a Feb. 2 email from Deanna Garrett-Ostermiller, assistant director of the Center for Student Conduct.

She believes the University needs to consider when AI tools are suitable for use in classes and assignments, and when use of the tools would constitute an academic integrity violation, she stated. She also believes it is important to teach students how to use the tools effectively and ethically.

There have been no reports of academic integrity violations related to ChatGPT, she stated. Sanctions for academic dishonesty using AI tools would be similar to sanctions for other methods of plagiarism and would vary based on the incident, she stated.

Some classes may include instruction to students about how they can use AI tools responsibly and ethically, she stated. As this AI technology evolves, additional education concerning AI tools may be needed in the future.

These AI tools give students the ability to answer questions using the AI, Cafiero said. He said he has seen ChatGPT write entire short programs, consisting of 10 to 50 lines of code, with high accuracy.

Another issue arises when AI sometimes offers what seems like a plausible solution to a beginner, but is actually completely wrong, Cafiero said. A student may submit the answer to a problem for an assignment, but would have no idea if it was correct or not.

Right now, the computer science department is relying on trust and is not currently using any measures to detect AI in student work, he said. Homework assignments submitted by students are auto-graded, but either the instructor or a teaching assistant look at all submissions before finalizing the grade.

Minor changes to the structure of some computer science courses, such as modifying the weighting of assignments, are being implemented as a result of emerging AI, Cafiero said. No other changes to his class are planned for now, he said.

The hope is that by making homework assignments carry lower stakes, students will be more encouraged to do the work themselves rather than use AI, Cafiero said. He has yet to catch any of his students using AI in their work.

AI tools could be useful for students in some contexts, Cafiero said. They could help students better understand concepts of programming by providing clear explanations and examples, allowing them to more easily apply them into their work.

Some educators are taking the position that it is not worth it to put in extra time to determine if students use AI tools because the consequences will present themselves to the students in the future, said Andrew Barnaby, an English professor.

Since COVID-19, HR vendors have been increasingly creating tools that attempt to analyze employee behavior. The latest is from HCM vendor Ceridian: its "burnout dashboard" is designed to identify employees who may be disengaging, lacking motivation or souring on the workplace.

The new dashboard utilizes the World Health Organization's classification of burnout as an occupational syndrome. The indicators include depletion of energy or exhaustion, mental distance from one's job, negativity or cynicism about the job, diminished professional efficacy, or a dwindling sense of professional competence.

"Their performance -- either self-perceived or externally perceived -- is low," said Brittany Schmaling, a data analyst at Ceridian, which reported $1.2 billion in revenue last year and is based in Minneapolis.

By aggregating data across its HCM platform, Ceridian evaluates an organization's "work energy," which it defines as how someone shows up for work, such as engaged, disengaged, overextended, ineffective or burned out. Employees who show a high degree of cynicism, for example, may be detaching from their organization or role. A key indicator of burnout is the frequency of sick leave usage. Schmaling said burnt-out employees are 50% to 60% more likely to take a sick day.

Some indicators of burnout include changes in work patterns, such as failing to complete tasks and meet deadlines in as timely of a manner as an employee used to, along with that employee's level of engagement. The tool can also analyze groups or departments to see if there's a pattern of overwork, she said.

Although the tools used in productivity monitoring or burnout analysis are intended to help employees, they could have the opposite effect, said Ridhi Shetty, a policy counsel with the Center for Democracy and Technology's privacy and data project in Washington D.C.

She said burnout tools could penalize underperforming workers rather than support those experiencing work fatigue. It could even put some workers at greater risk of injury if they feel pressure to work faster.

Employees should know when these tools are in use, Shetty said, but it's a question of "what kind of notice is sufficient to ensure workers understand what they are being subjected to." And once workers know, "what actions can they take?" she added.

Schmaling said the vendor has discussed the ethics of its burnout dashboard, and it is giving guidance to "help employers think about how to help and do what's best for their employees." It also believes that the employees should be notified of the tool's use.

Natalie Pierce, chair of law firm Gunderson Dettmer's employment and labor practice, said monitoring tools could be helpful for employers looking at workplace patterns that may, for instance, point to unmanageable workloads. But employers must also consider the risk of employee backlash.

Employers don't want to recruit the best talent only to see them leave because of trust and transparency issues. If employees don't like the use of monitoring tools, employers may ask whether the tool is worth using, she said.

For instance, if an employee with a few years of tenure isn't taking advantage of other opportunities in the workplace, such as advancement possibilities, stretch assignments and working on their career path, that could be a sign of low engagement and a reason to follow-up, said Kumar Ananthanarayana, executive director of product management at Phenom People. Another information source is the check-ins between managers and employees.

Ananthanarayana believes more extensive monitoring of employees could hurt morale. But showing employees their growth opportunities, for instance, may "help them be more engaged and reduce that burnout from that perspective."

And yet, when each group discussed the underlying purpose of the business,everyone fell into agreement. Everyone wanted to make the capital markets a moreefficient and effective economic tool. That felt like noble work and what IR, atits best, could achieve.

In recent decades, public service provision has become increasingly digitalized. However, while digitalization and artificial intelligence holds many promises, there is surprisingly little causal evidence on how it affects the employees who provide such services in the frontline. Based on cognitive and social psychological theories, we argue that IT projects can increase employees' cynicism towards change and change fatigue. In liaison with a Danish unemployment insurance fund, we test our hypotheses in a pre-registered randomized controlled trial that introduced an algorithmic decision-making support tool to underpin the counselling of newly unemployed clients. We do not find evidence that implementation of this tool resulted in negative employee outcomes. However, exploratory analyses indicate that this conclusion may mask smaller or heterogenous effects depending on employees' years of service with the insurance fund. We end the paper by discussing the implications of organizational change in the public sector.

N2 - In recent decades, public service provision has become increasingly digitalized. However, while digitalization and artificial intelligence holds many promises, there is surprisingly little causal evidence on how it affects the employees who provide such services in the frontline. Based on cognitive and social psychological theories, we argue that IT projects can increase employees' cynicism towards change and change fatigue. In liaison with a Danish unemployment insurance fund, we test our hypotheses in a pre-registered randomized controlled trial that introduced an algorithmic decision-making support tool to underpin the counselling of newly unemployed clients. We do not find evidence that implementation of this tool resulted in negative employee outcomes. However, exploratory analyses indicate that this conclusion may mask smaller or heterogenous effects depending on employees' years of service with the insurance fund. We end the paper by discussing the implications of organizational change in the public sector.

AB - In recent decades, public service provision has become increasingly digitalized. However, while digitalization and artificial intelligence holds many promises, there is surprisingly little causal evidence on how it affects the employees who provide such services in the frontline. Based on cognitive and social psychological theories, we argue that IT projects can increase employees' cynicism towards change and change fatigue. In liaison with a Danish unemployment insurance fund, we test our hypotheses in a pre-registered randomized controlled trial that introduced an algorithmic decision-making support tool to underpin the counselling of newly unemployed clients. We do not find evidence that implementation of this tool resulted in negative employee outcomes. However, exploratory analyses indicate that this conclusion may mask smaller or heterogenous effects depending on employees' years of service with the insurance fund. We end the paper by discussing the implications of organizational change in the public sector.

All content on this site: Copyright 2024 Elsevier B.V. or its licensors and contributors. All rights are reserved, including those for text and data mining, AI training, and similar technologies. For all open access content, the Creative Commons licensing terms apply

c80f0f1006