REPORT FOR RESEARCH:
HOW DOES THE INTEGRATION OF ARTIFICIAL INTELLIGENCE
TECHNOLOGIES IN THE U.S. TECHNOLOGY SECTOR AFFECT
EMPLOYMENT RATES AND WAGE INEQUALITY BETWEEN HIGH-
SKILLED AND LOW-SKILLED WORKERS?
1-) SELECTION OF THE RESEARCH QUESTION
In 2023, OpenAI launched ChatGTP and the GTP model, which created a leap forward in
arti cial intelligence. Arti cial intelligence, which has the potential to revolutionize many
elds with its power in automation, brings many advantages as well as disadvantages. As
with many technological and scienti c developments, innovation is also disruptive.
Automation reduces the productivity of companies and organizations and the need for
labor, while increasing unemployment and inequality. The development of arti cial
intelligence reduces or eliminates the need for labor in many elds and is leading to
signi cant change and transformation in labor markets. According to The Anthropic
Economic Index Report, the areas where AI will have the least impact are jobs that require
few quali cations, such as farming, or jobs that require very high quali cations and
manual dexterity, such as obstetrics [1]. The implication is that a certain level of
technological infrastructure is required for AI transformation. Technology that already
exists at a certain level can be automated and included in the transformative circle of AI.
Based on an analysis of nearly one million
conversations on Claude.ai, the conversations were
matched to tasks in the US Department of Labor's
O*NET database, and the dataset output (Graph-1)
shows that in the “Computer and Mathematical” eld,
the proportion of Claude conversations is much
higher than in the US workforce [1]. This explains why
arti cial intelligence, which enables technological
automation, has the highest use in the technology
sector. This provides insight into the possibility of an
exponential, rather than linear, change in
technological transformation. This means that the
positive and negative e ects of AI will be rst and
Graph-1: AI uses by industries in the US strongest in the technology sector. Being able to
tolerate this change and transformation, adaptability
and exibility are of paramount importance, but do companies and organizations really
have this adaptability and speed?
In the early 2000s, being able to use a computer could mean a great competence and
knowing excel could mean being able to use advanced technology. Every development,
especially in computer science, has rede ned the concept of “advanced technology”. The
development of arti cial intelligence means that yesterday's skilled people who can use
advanced technology can turn into tomorrow's mid-to-low-level technology workers; if
their work has not yet been automated.
According to the U.S. BUREAU OF LABOR STATISTICS data, the salary distribution in
selected elds in the technology industry for 2023 is as follows [2]. In this platform, the
data (Table-1) is available on the In Occupational Employment and Wage Statistics Query
System page:
fi
fi
fi
fl
fi
fi
fi
fi
ff
fi
fi
fi
fi
fi
fi
fi
fi
Multiple occupations for one industry > Professional, Scienti c, and Technical Services
(sector 54) > Computer systems design and related Services
Selected elds:
• Computer and Mathematical Occupations
• Engineers
• Installation, Maintenance, and Repair occupation
• Production occupations
Table-1: Salaries for workers in selected tech elds 2023 US Graph-2: Visualization for Table-1 data
Data for selected technology areas were analyzed and visualized using Python and data
libraries (pandas, matplotlib). Hourly wages in selected areas are more than double those
in some technology occupations.
In conclusion, the reason why I chose this research topic and its key points;
- Addressing the AI transformation in a general framework, staying away from sector-
speci c determinations to a certain extent
- Data-driven research on whether elds in the technology industry, such as software, are
indeed at risk of extinction after AI
- Investigating how the threats of inequality posed by arti cial intelligence will a ect the
technology sector involved in the production of arti cial intelligence
- Investigating the impact of the change in the de nition of the concept of “advanced
technology” and the increase in the rate of change on individuals and organizations that
delay in the use/integration of arti cial intelligence
- I made a comparison based on “salary” since this e ect is measurable and data can be
accessed e ectively.
- Again, I wanted to make this comparison within the US in terms of data accessibility and
the abundance of research.
2-) LITERATURE REVIEW
In order to keep my research sources diverse, I used various databases, research reports,
articles, datasets, and to reinforce my background knowledge on the subject, I used The
Narrow Corridor by Daron Acemoğlu & James A. Robinson: States, Societies, and the
Fate of Liberty [3].
While scanning the sources for my research topic, I used The Anthropic Economic Index
report I mentioned in part 1(Selection of research question). Here I used various
comparisons and data on the use of arti cial intelligence across sectors [1]. I used the
U.S. Bureau of Labor Statistics, which is a database for comparing salaries in the US
technology sector [2]. I downloaded the data set here and visualized the data using
python data science and visualization libraries (Pandas, Matplotlib, Seaborn) via Colab [4].
fi
fi
ff
fi
fi
fi
fi
fi
fi
ff
fi
fi
ff
2.1. OTHER ARTICLES AND RESOURCES
1. The impact of AI on the labour market: is this time di erent? (Marguerita Lane, 2021) [5]
✤ This source helps me to understand the impact of arti cial intelligence on the labor
market. Speci cally, its emphasis on how past technological changes didn't lead to
mass unemployment but caused transformations in the labor market allowed me to
see that AI could have a similar impact, though this transformation might be faster and
more comprehensive. These ndings helped me analyze how AI could increase wage
inequality between high-skilled and low-skilled workers in my research. Additionally, it
guided me in developing policy recommendations, such as the importance of reskilling
and training programs.
2. The Impact of Arti cial Intelligence on Employment and Income Distribution (Yihang
Liang 2024) [6]
✤ This article provides a comprehensive analysis of the impact of arti cial intelligence
(AI) on income distribution and employment. Speci cally, it explains how AI
exacerbates wage inequality between high-skilled and low-skilled workers through
mechanisms such as skill bias and capital bias. Additionally, the article discusses
how AI replaces low-skilled jobs while creating new high-skilled jobs, thereby
reshaping the employment structure. These insights are highly relevant to my research
question, as they provide both theoretical and empirical foundations for understanding
the e ects of AI on employment and wage inequality.
3. Tasks, Automation, and the Rise in US Wage Inequality ( Acemoglu and Restrepo,
2021) [7]
✤ This study is essential for my research as it sheds light on how automation impacts
wage structures, task displacement, wage changes, and broader economic e ects. It
provides strong evidence that 50% to 70% of changes in the U.S. wage structure over
the past four decades can be linked to the wage declines of workers focused on
routine tasks, especially in industries rapidly adopting automation. The study’s
framework connects task displacement to wage changes, o ering a clear explanation
for rising wage inequality. Additionally, it introduces a method to assess the wider
economic impacts of automation, making it a valuable resource for understanding how
automation shapes labor markets and wage dynamics. What makes this study
particularly useful is its ability to separate the e ects of automation from other
in uences like declining unionization and market power changes
4. The Future of Jobs Report 2025 - World Economic Forum (WEF) [8]
✤ This report is a critical resource for understanding the future trends in the labor
market, particularly the impact of automation, arti cial intelligence, and digital
transformation on employment and skill requirements. I am including this report in my
research because it provides a comprehensive analysis of how automation will a ect
jobs across di erent sectors, the emergence of new roles, and the skills that will be in
demand in the future. The report o ers valuable insights for policymakers, businesses,
and individuals by presenting a global perspective on labor market changes.
Therefore, I will use this report to analyze the e ects of automation on employment
and to explore the dynamics of the future labor market.
fl
ff
fi
ff
fi
fi
ff
ff
ff
fi
fi
fi
ff
ff
fi
ff
ff
5. 2024 AI Index Report- Stanford University, Human-Centered Arti cial Intelligence [9]
✤ This report provides a comprehensive analysis
of the latest developments, trends, and impacts
in the eld of arti cial intelligence. The report
examines AI's applications across di erent
sectors, its impact on the workforce, and its
future potential. The section that speci cally
analyzes labor markets in the context of AI is
particularly relevant to my research. This part
delves into how AI automates jobs for workers in
routine tasks while creating new opportunities in high-skill roles. Therefore, I am using
this report to analyze the e ects of AI on the labor market.
6. Future of Jobs Report 2025 by the World Economic Forum [10]
I utilized this report in my research because it provides a comprehensive analysis of how
arti cial intelligence impacts labor markets. The report contains particularly valuable data
demonstrating how AI adoption in the U.S. technology sector has exacerbated wage
inequality between high-skilled and low-skilled workers. Its global employer surveys and
employment projections serve as reliable evidence supporting my ndings. Additionally,
the report o ers meaningful policy recommendations for mitigating the negative
employment e ects of technological transformation, making it an authoritative reference
for my study.
3-) AI TOOLS I USED
DeepSeek [11]: I used it to summarize sources and articles, to determine whether
they would provide insights into my research topic.
Semantic Scholar [12]: A research tool supported by arti cial intelligence. It provided
the opportunity to chat with articles. It allowed me to extract the maximum information
I could from an article for my research topic by communicating in a question and
answer format.
Elicit [13]: An AI-powered research tool. Takes the Research question as input and
gathers the most important studies, academic articles, reports and data in the eld
and presents them in a report. I used it in order to be able to see and analyze the
studies in the eld holistically.
Conected papers [14]: It is an AI tool that presents articles, reports and other sources
with a map at the contextual relationship level. It allowed me to expand my research
circle as much as possible, while adhering to my research topic.
Cursor [15]: An AI-powered code editor. It allowed me to improve my code for
analysis and visualization while examining salary data and to handle the data in a
multidimensional way with more diverse graphics.
4- ) CONCLUSION
I selected my research question by considering current technological advancements and
key economic challenges and threats. To ensure data accessibility and availability, I
focused on the United States as the primary case study. For measurability, I chose
"wages" as the central metric to analyze the impact of AI on employment and inequality.
fi
fi
ff
fi
ff
fi
ff
ff
fi
fi
fi
fi
fi
In conducting the literature review, I prioritized diversity in sources, incorporating not only
datasets but also academic studies and reports to provide a comprehensive perspective.
To enhance e ciency and maintain a systematic research process, I leveraged various AI-
powered tools. Additionally, I continue to engage with relevant literature to deepen my
background knowledge and fully internalize the subject matter.
REFERENCES
[1] Anthropic. (2024). The Anthropic Economic Index. Anthropic. https://www.anthropic.com/
news/the-anthropic-economic-index
[2] U.S. Bureau of Labor Statistics. (2024). Occupational Outlook Handbook, from https://
www.bls.gov/ooh/
[3] Acemoğlu, D., & Robinson, J. A. (2019). The narrow corridor: States, societies, and the fate of
liberty. Penguin Press.
[4] https://colab.research.google.com/drive/1Qks9np9fX_rq1juvUHrUAcqDD9qDz-52?
usp=sharing
[5] Lane, M. (2021, January 25). The impact of AI on the labour market: Is this time
di erent? OECD AI Policy Observatory. https://oecd.ai/
[6] Liu, Y. (2024). Exploring The Impact Path of Arti cial Intelligence Development on Income
Distribution Equity. Journal of Education, Humanities and Social Sciences.
[7] Acemoglu, D., & Restrepo, P. (2021). Tasks, Automation, and the Rise in US Wage
Inequality. SSRN Electronic Journal.
[8] World Economic Forum. (2025). The Future of Jobs Report 2025. World Economic
Forum. https://www.weforum.org/publications/the-future-of-jobs-report-2025/
[9] Stanford University, Human-Centered Arti cial Intelligence (HAI). (2024). 2024 AI Index
Report. Stanford HAI. https://hai.stanford.edu/ai-index/2024-ai-index-report/
[10] https://www.weforum.org/publications/the-future-of-jobs-report-2025/
[11] DeepSeek. (2025). DeepSeek Chat (Version 2025.03) [Large language model]. https://
www.deepseek.com
[12] Allen Institute for AI. (2025). Semantic Scholar [AI research tool]. https://
www.semanticscholar.org
[13] Ought. (2025). Elicit: The AI Research Assistant (Version 2.5) [Software]. https://elicit.org
[14] Connected Papers. (2025). Connected Papers [Literature visualization tool]. https://
www.connectedpapers.com
[15] Anysphere. (2025). Cursor AI (Version 1.8) [AI-assisted code editor]. https://www.cursor.sh
ff
ffi
fi
fi
