The Impact of AI on Business Students’ Career Choices

Zihang Gong

Business Analytics, School of Business, Macau University of Science and Technology,

Macao Special Administrative Region, 999078, China

Keywords: AI, Business, Salary, Job, Educational Level.

Abstract: The growing popularity and evolving nature of artificial intelligence (AI) technology have significantly

influenced numerous industries and sectors in contemporary society. Concurrently, there has been a noticeable

shift towards AI-related majors in higher education. Given the diverse range of business disciplines, the

question of whether business students should acquire AI skills in their career choices merits thorough

examination. This study relies on salary report data compiled by Kaggle, encompassing 6,704 samples from

surveys, job boards, and other public sources. Utilizing regression analysis and data virtualization techniques,

the study objectively assesses the impact of AI technology on business operations and its subsequent effect

on compensation. The findings reveal that, aside from master's degree holders, there exists a notable

association between the salary of undergraduate and doctoral graduates and the extent of AI involvement in

their work. Furthermore, the study identifies a correlation between educational level and salary. Additionally,

across all business-related professions, personal work experience is found to have a greater influence on salary

than educational qualifications.

1 INTRODUCTION

The present situation of social employment is

experiencing a marked deterioration. The COVID-19

pandemic has exerted considerable influence on the

conduct of businesses, leading to a substantial decline

in recruitment demand. Additionally, numerous small

and micro-enterprises are confronted with a critical

survival challenge. As the primary providers of

employment opportunities for graduates, the

struggling state of small, medium, and micro-

enterprises has had a profound impact on the

employment landscape. The Employment Market

Sentiment Index (CIER Index) serves as a crucial

metric to gauge the state of the job market. According

to the "China Job Market Sentiment Report" issued

by Zhaopin Ltd, the increase in the number of job

applicants in the first quarter of 2022 surpassed the

growth in recruitment needs, resulting in a decline in

the CIER index. This indicates that the present

employment situation is not optimistic (Yan,2022).

Simultaneously, the widespread utilization of

digital technologies, including artificial intelligence,

big data, and cloud computing, has exerted a

https://orcid.org/0009-0002-6563-5800

significant influence on diverse industries (Chen,

2019). Taking accounting work as an illustrative

example, certain conventional roles such as

fundamental accounting and cashiers might undergo

gradual disappearance, whereas others will

adaptively evolve alongside alterations in market

demand. In order to acclimatize to the perpetual

advancement of enterprise technology and business

models, novel, comprehensive accounting positions

will arise as necessitated by the evolving era (Wu,

2023).

College graduates, being a crucial segment in the

job market, possess employment quality that is a

fundamental prerequisite for the superior

advancement of higher education. In the present

challenging employment landscape, the employment

difficulties faced by university students have become

increasingly apparent. Being a susceptible segment of

the workforce, their employment situation deserves

greater consideration. Statistics indicate that in 2019,

the number of individuals supported by various

education funding schemes throughout the country

reached 106 million, an increase of 8.05% year-on-

168

Gong, Z.

The Impact of AI on Business Students’ Career Choices.

DOI: 10.5220/0012917700004508

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Engineering Management, Information Technology and Intelligence (EMITI 2024), pages 168-174

ISBN: 978-989-758-713-9

year, with a total allocation of 212.6 billion yuan

(Chen, 2022).

A preliminary examination of employment

recommendations and adaptive strategies across

various business sectors has been undertaken.

Notably, sectors like finance and accounting have

garnered significant attention due to their extensive

workforce, long-standing history, and traditional

stronghold status within the business world.

Nevertheless, these industries also encounter

challenges such as limited salary growth, rigid job

responsibilities, and industry saturation. The evolving

economic landscape has given birth to novel

industries. Graduates born in the early 2000s have

demonstrated a keen interest in emerging professions,

attaching greater importance to salary scales and

industry outlooks. According to a survey conducted

by the United Questionnaire Network of the Social

Survey Center of China Youth Daily, over 90% of the

surveyed graduates born in the early 2000s expressed

an interest in pursuing new careers, with more than

30% expressing considerable enthusiasm.

Furthermore, their job search preferences have

become increasingly diverse (Yan, 2022).

As such, the research will zero in on the evolving

business landscapes and investigate their integration

with AI technology. Through comparing various

business roles, it strives to offer an accurate portrayal

of the diverse professional landscapes prevailing in

modern business schools, thus providing young

individuals with ample references for pursuing either

traditional or emerging career paths (Wei, 2023).

This article is grounded in Kaggle's salary report

data, alongside surveys, recruitment platforms, and

various public resources, to delve into the precise

influence of AI on salaries within the business sector.

With the ultimate goal of achieving optimal

remuneration under identical circumstances, this

research evaluates the necessity of AI proficiency

among business students based on the analytical

outcomes. Furthermore, the author considers multiple

variables, including work experience and educational

attainment, to offer tailored guidance for personal

development among diverse student groups. This

approach aims to equip them with the necessary skills

to thrive in the age of AI.

2 METHOD

The primary source of data employed in this article

originates from the Salary_Data dataset, which was

procured from the Kaggle website. This dataset

encompasses a diverse array of sources,

encompassing surveys, recruitment websites, and

other publicly accessible resources. A comprehensive

collection of 6704 data points was compiled. The

dataset encompasses six variables: age, gender,

experience, job, education, and salary (Alekseeva,

2021). While these variables are rich in content, age,

and gender are deemed irrelevant in this study due to

the authors' focus on the broader context of the

research. These two variables are particularly delicate

when addressing social audience groups and are often

influenced by geography or specific job restrictions

without further reference.

Using Excel-based skill data processing

techniques, including regression analysis and data

virtualization, the author meticulously sifted through

6704 data points (Seamans, 2018). Recognizing the

vast array of business-related professions, a focused

selection of six mid-level positions was made,

drawing from a comprehensive survey of business

school graduates: Data Analyst, Digital Marketing

Manager, Human Analyst, Marketing Manager,

Operations Manager, and Product Manager. Notably,

three of these roles - Data Analyst, Digital Marketing

Manager, and Marketing Analyst - exhibit a high

degree of integration with the AI industry, hence their

classification as highly relevant. On the other hand,

Operations Manager and Product Manager are

deemed moderately related, while Human Resource

Manager and Marketing Manager are considered less

so. Following this filtering process, 1333 data items

that met the established criteria remained.

Subsequently, two non-pertinent variables, Age and

Gender, were discarded. The refined dataset now

comprises three independent variables: Education

Level, Job Title, and Years of Experience, and the

dependent variable: is salary.

2.1 Data Processing

2.1.1 Group A: Relevance of Job Wages to

Jobs and AI

This study aims to investigate the influence of

varying positions on salary scales. To ensure the

precision of the findings, education level is

introduced as an additional crucial variable. The

educational level is categorized into three tiers:

Bachelor's, Master's, and PhD. The data for each of

these tiers is processed independently. To uphold the

study's rigor, a confidence interval of 95% has been

established.

In Table 1, solely undergraduate-level data is

presented. Within this table, data is further segmented

into high, medium, and low levels, and processed

The Impact of AI on Business Students’ Career Choices

169

accordingly. To facilitate statistical analysis, dummy

coding is employed, assigning numerical values of 1,

2, and 3 to high, medium, and low levels,

respectively. The data in Table 2 exclusively pertains

to the master's level, and the processing approach

mirrors that of Table 1. Similarly, the data in Table 3

solely concerns the doctoral level, adhering to the

same processing methodologies as the preceding

tables.

2.1.2 Group B: The Relationship Between

Education Level and Salary

Unlike the processing approach employed by Group

A, the author has categorized the data into three

distinct groups: High, Medium, and Low, based

solely on the Job Title. Each of these sheets has been

assigned a confidence interval of 95%. By adjusting

for job type, an independent and thorough analysis of

educational attainment can be conducted.

Table 4 presents a detailed breakdown of the

High-level data, encompassing the four distinct

education levels: High School, Bachelor, Master, and

PhD. To facilitate analysis, virtual data values have

been assigned to each level: 0, 1, 2, and 3,

respectively. Table 5 showcases Medium-level data,

which includes three education levels: Bachelor,

Master, and PhD. For these levels, the virtual data

assignments are 1, 2, and 3. Table 6 solely focuses on

PhD-level data, while still incorporating three

education levels: Bachelor, Master, and PhD.

Similarly, the virtual data assignments for these levels

are 1, 2, and 3.

2.1.3 Group C: Work Experience or

Education, Which One Has a Greater

Impact on Salary?

In addressing this problem, given the involvement of

multiple variables, the author deemed it necessary to

employ the virtual data processing method again,

building upon the foundation established by Group B.

Specifically, the objective is to determine the mean

value of "Years of Experience" across three distinct

data categories: High, Medium, and Low.

Subsequently, any data point with a mean value equal

to or less than the computed mean will be assigned a

value of 0, while those exceeding the mean will be

assigned a value of 1. Notably, Variable 1 signifies

education level, while Variable 2 represents work

experience.

To assess which factor exerts a more significant

influence on salary, a comparison of the P values

across various groups is essential. In ensuring the

reliability of the results, the authors have established

a confidence interval of 95%. Referring to the data

presented in Table 7, the average "Year of

Experience" for the High level is 4.75. Consequently,

any data point with a value less than or equal to 4.75

will be set to 0, while those exceeding 4.75 will be set

to 1. Similarly, Table 8 reveals an average "Year of

Experience" of 7.85 for the Medium level, resulting

in a similar binary classification. Finally, Table 9

indicates an average value of 9.55 for the Low level,

leading to the same binary assignment of data points.

2.2 Experimental Hypothesis

Group A: Education level directly correlates with the

job salary's relation to AI. As education rises, so does

its connection with AI-related salaries.

Group B: AI-reliant jobs show a positive correlation

between salary and varying education levels. Higher

education is linked to stronger salary correlations,

while lower education is linked to weaker ones.

Group C: When considering both education and work

experience's impact on salary, work experience has a

greater influence than education.

3 RESULT

3.1 Data Results

Table 1: Regression Statistics of Bachelor’s Level.

Regression

Statistics

Multiple R

0.108973

Observations 800

Coefficie

nts

Standard

Erro

t Stat P-value

Intercept

115541.7

866

3401.052

362

33.97236

335

3.5603E

-157

Panel Data

-6182.

235018

1996.306

423

-3.0968

36711

0.00202

442

Table 2: Regression Statistics of Master’s Level.

Regression

Statistics

Multiple R

0.070651

773

Observations 489

Coefficie

nts

Standard

Erro

t Stat P-value

Intercept

128990.2

249

5178.532

452

24.9086

447

6.37301

E-89

Panel Data

3353.428

839

2145.432

756

1.56305

474

0.118689

274

EMITI 2024 - International Conference on Engineering Management, Information Technology and Intelligence

170

Table 3: Regression Statistics of PhD Level.

Regression

Statistics

Multiple R

0.461040

894

Observations 37

Coefficie

nts

Standard

Erro

t Stat P-value

Intercept

237887.3

239

48651.95

594

4.889573

695

2.23849

E-05

Panel Data

-52517.

60563

17086.00

-3.0737

20533

0.00408

0626

Table 4: Regression Statistics of High-level Data.

Regression

Statistics

Multiple R

0.04387

0473

Observations 547

Coefficie

nts

Standard

Erro

t Stat P-value

Intercept

102188.

6636

5037.271

112

20.2865

1254

1.386E-

Panel Data

4152.33

8037

4050.454

961

1.02515

3489

0.30574

5563

Table 5: Regression Statistics of Medium-level Data.

Regression

Statistics

Multiple R

0.60814

8459

Observations 427

Coefficie

nts

Standard

Erro

t Stat P-value

Intercept

52163.0

9322

5256.446

047

9.92364

2847

5.15933

E-21

Penal Data

56465.8

0508

3575.240

803

15.7935

6698

1.52897

E-44

Table 6: Regression Statistics of Low-level Data.

Regression

Statistics

Multiple R

0.14643

1147

Observations 359

Coefficie

nts

Standard

Erro

t Stat P-value

Intercept

98218.3

0422

6145.214

753

15.9828

9208

9.38245

E-44

Years of

Experience

8959.98

2982

3203.560

2.79688

3036

0.00543

9286

Table 7: Regression Statistics of High-level Data.

Regression

Statistics

Multiple R

0.681154

913

Observations 547

Coefficie

nts

Standard

Erro

t Stat P-value

Intercept

111081.0

3717.679

256

29.87914

916

8.1448E

-117

Va r i a ble 1

-22293.

70502

3212.293

588

-6.9401

20638

1.11855

E-11

Va r i a ble 2

59093.66

2728.927

027

21.65454

1.84851

E-75

Table 8: Regression Statistics of Medium-level Data.

Regression

Statistics

Multiple R

0.81616

2725

Observations 427

Coefficie

nts

Standard

Erro

t Stat P-value

Intercept

52658.6

6536

3830.749

238

13.7463

0969

8.0736E-

Va r i a ble 1

37604.9

1234

2781.003

443

13.5220

6608

6.83981

E-35

Va r i a ble 2

55037.6

5919

2837.410

618

19.3971

4289

1.85588

E-60

Table 9: Regression Statistics of Low-level Data.

Regression

Statistics

Multiple R

0.588060

observations 359

Coefficie

nts

Standard

Erro

t Stat P-value

Intercept

98428.26

418

5031.590

994

19.56205

588

2.19329

E-58

Va r i a ble 1

-486.17

68246

2717.654

546

-0.1788

95741

0.85812

1333

Va r i a ble 2

40726.11

713

3065.321

143

13.28608

496

5.41459

E-33

3.2 Experimental Results

The confidence intervals of the above charts are all

set to 95%. With the above data, you researchers can

find:

Group A: The results of Table 1 demonstrate

statistical significance, albeit with a weak correlation.

Conversely, the findings presented in Table 2 fail to

exhibit significance, and the correlation is weak.

The Impact of AI on Business Students’ Career Choices

171

Lastly, Table 3 shows both statistical significance and

a strong correlation.

After careful analysis, it is evident that the salary

of individuals holding a bachelor's degree

demonstrates a limited association with the AI sector.

In contrast, graduates with a master's degree exhibit

negligible correlation with the AI industry when it

comes to salary. However, a strong link is observed

between the salary of PhD graduates and the AI field.

Group B: The statistical significance of the data

presented in Table 4 is lacking, indicating a non-

significant correlation. Similarly, while the findings

in Table 5 and Table 6 demonstrate statistical

significance, they do not exhibit a strong correlation.

It is evident that in professions that are heavily

reliant on AI, the influence of educational attainment

on remuneration is relatively minor. Conversely, in

professions that have a moderate association with AI,

the link between salary and educational level is more

pronounced, typically manifesting as higher levels of

education leading to higher salaries. In professions

that have a tenuous connection to AI, a certain degree

of correlation exists between educational attainment

and salary, albeit not a strong one.

Group C: Tables 7, 8, and 9 indicate that Variable

2 is smaller across all job types. Specifically, for high-

level job types, work experience has a more

significant influence on salary compared to

educational experience. Similarly, for medium-level

job types, the impact of work experience on salary

surpasses the influence of educational experience.

Lastly, for low-level job types, work experience

remains the primary factor determining salary,

outweighing the influence of educational experience.

Judging from these three situations, no matter

what business-related occupation you are engaged in,

the impact of work experience on salary is generally

greater than the impact of educational experience on

salary.

After a thorough examination, the disparities

observed between Groups A and B can be attributed

primarily to the varying levels of professional

proficiency among their members. Research

conducted by Fang Ning and colleagues has revealed

that the marketplace holds a significant demand for

undergraduate students. Specifically, 66.39% of

employers express a preference for undergraduate

students over master's degree holders. Conversely,

only 13.68% of employers prefer master's degree

holders over undergraduate students. This

recruitment preference is primarily determined by the

extent to which an individual's skills align with the

requirements of the job, accounting for 60.05% of the

decision-making process (Fang, 2020).

It is noteworthy that employers highly regard the

diverse abilities of undergraduate and postgraduate

students, encompassing self-motivation, values, self-

confidence, oral communication skills, problem-

solving proficiency, and team collaboration abilities.

Notably, 55% of companies maintain that the work

competency of master's students in similar positions

does not significantly surpass that of undergraduates.

Consequently, the majority of employers do not

endorse the notion that only graduates with master's

degrees can secure superior employment

opportunities (Fang, 2020).

In terms of the specific perspectives held by

enterprises, it is noteworthy that 42.04% of

enterprises emphasize the importance of not solely

focusing on academic qualifications, but rather giving

greater consideration to personal abilities.

Furthermore, 4.8% of enterprises believe that

undergraduates possess considerable potential and

room for development once they have undergone

systematic training. Additionally, 43.38% of

enterprises maintain that, apart from certain

specialized technical positions, there is not a

significant difference in quality between master's

students and undergraduates. Notably, only 9.7% of

companies hold the view that academic qualifications

are of utmost importance (Fang, 2020).

From the perspective of talent demand within the

AI industry, employers place a greater emphasis on

candidates' personal abilities than on their academic

qualifications. When evaluating candidates' abilities,

employers prioritize professionalism, followed by

professional ability. Notably, demands for

professional ethics, self-motivation, initiative, and

collaboration skills are particularly acute. These

abilities are primarily shaped by an individual's

fundamental character traits, perseverance, and work

ethic, rather than being directly influenced by

academic qualifications. While master's degree

holders may possess superior professional abilities,

these skills are relatively less in demand among

employers.

From the employers' perspective on evaluation

criteria, undergraduate students exhibit a high level of

proficiency in work-related abilities. Notably, there is

no significant difference in performance between

undergraduate and master's degree students,

indicating that academic qualifications do not solely

determine an individual's practical capabilities.

Indeed, over half of employers maintain that the work

capabilities of undergraduate students are comparable

to those of master's degree holders.

EMITI 2024 - International Conference on Engineering Management, Information Technology and Intelligence

172

3.3 Personal Development Suggestions

For novice students entering the realm of business,

selecting a major serves as a pivotal junction.

Whether they opt for business analysis, which holds

a strong nexus with AI technology, or major in

finance and human resources management, which

maintain a more tenuous tie with AI technology, it is

imperative for students to ascertain their career

aspirations and craft corresponding plans for

academic advancement and professional

development.

Firstly, those who opt for business analytics as

their field of study are often drawn to it due to its close

association with the prevalent AI technology. In this

domain, the significance of personal practical

proficiency is paramount (Liu, 2023). This

necessitates, alongside theoretical knowledge, the

mastery of practical analytical tools and techniques,

enabling students to leverage data in decision-making

and generate value for enterprises. This proficiency is

not solely contingent upon academic qualifications

but also hinges upon practical experience and skill

enhancement. The diverse resources and avenues

offered by the institution, including internships and

project-based practice, furnish students with an ideal

platform to hone these abilities. Consequently, for

students pursuing business analysis as their major, the

utilization of these resources and the accumulation of

practical work experience hold the utmost importance

in standing out during the job search process.

Certainly, this is not to suggest that the

significance of academic credentials be disregarded

by those pursuing a business analysis major. Indeed,

within the realm of business analysis, the possession

of an advanced degree frequently offers students a

wider array of development opportunities.

Nevertheless, alongside academic credentials, an

individual's practical abilities hold equal significance.

During the job search, companies tend to prioritize

the actual skills and potential of candidates,

surpassing mere academic qualifications. Therefore,

alongside the pursuit of academic credentials,

students majoring in business analysis must also

prioritize the enhancement of their own abilities.

On the contrary, students pursuing majors such as

finance or human resource management may not

possess the same level of specialization as those

majoring in business analysis. Nonetheless, this does

not negate the significance of academic qualifications

for these students (Liu, 2023). Quite the contrary,

enhancing their academic credentials often affords

them a competitive edge. This is attributed to the fact

that these majors demand a profound academic

foundation and knowledge reserve. For instance,

finance majors must master intricate financial

theories and analytical tools, whereas human resource

management majors must comprehend organizational

behavior, labor laws, and pertinent knowledge. By

persevering in their studies, these students can not

only acquire more structured academic training but

also elevate their professional qualities and

comprehensive abilities.

Furthermore, as technology continues to evolve,

AI technology is increasingly infiltrating sectors like

finance and human resources management.

Consequently, it's crucial for students pursuing these

majors to acquire a fundamental understanding and

proficiency in AI technology (Huang, 2023). While

they might not directly utilize AI technology in the

same manner as business analysis majors,

comprehending AI's applications, potential, and

integration with their respective fields will

undoubtedly afford them a competitive edge in their

future professional pursuits.

Therefore, for novice students venturing into the

realm of business, regardless of their chosen

discipline, it is imperative to clarify their career

aspirations and devise corresponding plans for

academic pursuits and professional development. In

sectors that prioritize artificial intelligence, personal,

hands-on capabilities hold the utmost significance;

whereas in domains like finance and human resources

management, the enhancement of academic

credentials and professional attributes is equally

crucial (Li, 2023). Furthermore, comprehending and

proficiency in AI technology stand as pivotal factors

for future career advancement. By leveraging the

resources and opportunities offered by the institution,

and consistently enhancing their practical skills and

professional attributes, students in the field of

business can secure a rewarding position upon

graduation and realize their career aspirations.

4 CONCLUSIONS

The research conducted by Group A has identified a

potential trend: there appears to be no substantial

direct link between the salary offered to master's

degree graduates and their employment in the AI

industry. This observation can potentially be

attributed to the diverse factors that influence salary

levels in the AI industry, encompassing work

The Impact of AI on Business Students’ Career Choices

173

experience, proficiency in skills, and the size of the

employing company, among others. Consequently,

when pursuing a career path, master's degree

graduates should not solely prioritize the higher

salaries offered in the AI industry, but rather, they

should holistically consider various pertinent factors.

The research undertaken by Group B has also

uncovered an intriguing trend: in professions that are

closely aligned with AI, disparities in educational

attainment do not significantly impact salary

outcomes. This could be due to the fact that in the

realm of AI, proficiency in skills and accumulated

experience are deemed more crucial than academic

qualifications. Regardless of one's educational level,

as long as one possesses the requisite skills and

experience, they are likely to be offered comparable

salary levels. This revelation holds significant

implications for both the education sector and

individuals seeking employment. The education

sector, particularly, ought to prioritize the cultivation

of practical abilities and skills, while job seekers need

to place a stronger emphasis on skill development and

the accumulation of practical experience.

The experimental findings of Group C align with

the widespread societal recognition that work

experience holds greater significance than

educational attainment. This finding echoes the

results obtained by Groups A and B, further

highlighting the importance of practical experience

and skills in career advancement.

This experiment relied primarily on limited real-

world data, which is accurate and representative but

may be incomplete. Extreme values or insufficient

data could influence the conclusion and deviate from

the actual situation. Extreme values may be caused by

errors or abnormal events, and if not handled properly,

they could significantly impact the conclusions. For

instance, limited data on doctoral job searches may

lead to overly optimistic conclusions. Ignoring other

factors, such as educational attainment, could also

lead to biased or one-sided conclusions. Despite these

limitations, careful design and in-depth analysis

allowed the experiment to draw valuable conclusions.

Future research will aim to improve methods and

techniques for more accurate conclusions.

REFERENCES

Chun, C., 2019. The promotion and impact of artificial

intelligence on the financial industry. Hebei Enterprise,

7, 60-61.

Chen, C., 2022. Research on the impact of financial aid on

the employment quality of college students with

difficulties—taking J University as an example.

Jiangxi: East China. Jiaotong University.

Hang, W., 2023. ChatGPT’s challenges and opportunities

for business. Foreign Economics and Management, 45,

9, 69-78.

Liudmila, A., et al. 2021. The demand for AI skills in the

labor market. Labour economics, 71.

Mian, Y., Fengcheng, L., Dong, Y., 2022. Problems and

countermeasures in the employment of college students.

Human Resources Development, 23, 61-63.

Maoyong, H., 2023. Asymmetric empowerment: the

iterative effect of artificial intelligence on employment

skills and educational intervention. Education and

Occupation, 1028, 4, 5-12.

Mingyu, L., 2023. Analysis and practical exploration of the

causes of college students’ employment difficulties

under the new situation. Economic Research Guide, 16,

135-137.

Ning, F, 2020. Comparative study on the employment

competitiveness of undergraduates and master's

students - based on the two dimensions of student

cognition and employer evaluation. Journal of Liaoning

Radio and Television University, 4, 87- 90.

Robert; S, 2018. AI, labor, productivity and the need for

firm-level data. National Bureau of Economic Research.

Wen, W., 2023. Duan Research on the training of

accounting talents in applied undergraduate colleges

under the background of artificial intelligence. Foreign

Economic and Trade, 5, 100-103.

Yongmou, L., 2023. The impact of "AI unemployment" on

contemporary youth and its response. Youth

Exploration, 1, 43-51.

EMITI 2024 - International Conference on Engineering Management, Information Technology and Intelligence

174