Future of Jobs and Generative AI

The advent of large language models (LLMs) like ChatGPT promises to transform the workplace by automating or augmenting a wide range of occupational tasks. However, a single perspective cannot fully grasp both the opportunities and risks these technologies represent across industries, workers, businesses and society. This article analyzes the World Economic Forum’s recent white paper [1] assessing the impact of LLMs on jobs through the lens of Spiral Dynamics. This integral framework reveals how different value systems perceive threats and opportunities differently. Administrative roles face disruption but efficiency gains (Blue). Innovative businesses are pressured to adopt but see new revenue potential (Orange). Vulnerable workers require support amidst job transformations (Green). Policymakers struggle to holistically analyze systemic impacts (Yellow). Realizing the benefits of LLMs requires honoring multiple worldviews, evolving processes, encouraging innovation, caring for people and conducting systems analysis. The analysis provides insights into LLMs’ multi-dimensional impacts and underscores the need for inclusive dialogue and initiatives to shape the AI-enabled future of work.


Here are the key points:

  1. LLMs could significantly impact many jobs due to their ability to automate or augment language-based tasks, which account for an estimated 62% of work time.
  2. The analysis assessed over 19,000 work tasks across 867 occupations to assess their LLM exposure. Tasks with high automation potential are routine and repetitive clerical/administrative tasks. Tasks with high augmentation potential require more abstract reasoning and problem-solving. Tasks with lower exposure potential emphasize interpersonal interaction.
  3. Occupations with the highest automation potential include credit authorizers, telemarketers, statistical assistants, and tellers. Occupations with the highest augmentation potential include insurance underwriters, bioengineers, mathematicians, and editors. Occupations with lower exposure include counselors, clergy, home health aides, and lawyers.
  4. Adopting LLMs will also likely create new roles like AI developers, content creators, interface designers, data curators, and AI ethics specialists.
  5. The financial services and information technology industries have the overall highest potential exposure. The finance and IT functional areas also have increased exposure.
  6. Significant alignment exists between occupations this analysis identifies as having high augmentation potential and those the Future of Jobs Report found to have high expected job growth. Similarly, occupations with high automation potential align with declining occupations.
  7. The report concludes LLMs will transform jobs and tasks, requiring strategies by businesses and government to prepare workforces for the change through training, transition support, and social safety nets. Overall, LLMs present opportunities to raise productivity and create new jobs, if managed responsibly.



Spiral Dynamics stages



What color are you Spiral Dynamics?


ColorBeigePurpleRedBlueOrangeGreenYellowTurquoise
In a lifeSurvivalFamily relationsThe rule of forceThe power of truthCompetitionInterpersonal relationsFlexible streamThe Global vision
In a businessOwn farmFamily businessStarting up a personal businessBusiness Process ManagementProject managementSocial networksWin-Win-Win behaviorSynthesis

Here is an analysis of the World Economic Forum white paper on large language models and jobs through the lens of Spiral Dynamics stages:


Spiral Dynamics StageQuotes from Document
 Beige No relevant quotes
 Purple No relevant quotes
 Red No relevant quotes
 Blue "With 62% of total work time involving language-based tasks, the widespread adoption of LLMs, such as ChatGPT, could significantly impact a broad spectrum of job roles." (p.4) This reflects the blue focus on structure, process and order.
 Orange "Adopting LLMs will transform business and the nature of work, displacing some existing jobs, enhancing others and ultimately creating many new roles." (p.19) This reflects the orange drive for innovation and progress.
 Green "Governments can also partner with and support employers and educational institutions to provide training programs that prepare workers for the jobs that will grow and benefit the most from LLMs. Additionally, social safety nets and assistance in transitioning to new roles will need to be reimagined and be more precisely targeted for those most likely to be affected." (p.19) This reflects the green concern for people and relationships.
 Yellow "To assess the impact of LLMs on jobs, this paper provides an analysis of over 19,000 individual tasks across 867 occupations, assessing the potential exposure of each task to LLM adoption, classifying them as tasks that have a high potential for automation, high potential for augmentation, low potential for either or are unaffected (non-language tasks). The paper also provides an overview of new roles that are emerging due to the adoption of LLMs." (p.4) This reflects yellow's emphasis on complex systems analysis.
 Turquoise No relevant quotes


The document overall reflects blue, orange, and green worldviews, with some elements of yellow systems thinking. There are no clear expressions of the beige, purple, red or turquoise value systems. This analysis illustrates how technology impacts different aspects of society and values.



Threats



Here is an analysis of threats and affected stakeholders through the lens of Spiral Dynamics stages:


Spiral Dynamics StageThreatsAffected Stakeholders
 Beige No major threats identified N/A
 Purple No major threats identified N/A
 Red No major threats identified N/A
 Blue Disruption of administrative processes and routines Organizations, administrative staff
 Orange Pressure to rapidly adopt new technologies Businesses, managers
 Green Job losses, inequality, lack of support during transition Individual workers, marginalized groups, society
 Yellow Complexity of analyzing and managing impacts Policy-makers, business leaders
 Turquoise No major threats identified N/A


In summary, the blue stage is threatened by disruption of established administrative processes, the orange faces pressure to innovate, the green risks job losses and inequality, and the yellow struggles with complex systems analysis. This highlights how different worldviews perceive threats and opportunities from the same technology trend. A holistic perspective is needed to understand the range of stakeholders and design responsible policies.


Elon Musk said about the danger of artificial intelligence (A.I.) in an interview with Tucker Carlson in April 2023. Below you can read an abridged version of the results of our VUCA poll "A.I. and the end of civilization". The full version of the results is available for free in the FAQ section after login or registration.

人工智能和文明的终结

國家
語言
-
Mail
重新计算
所有问题
所有问题
1) 安全(您同意或不同意多少?)
2) 控制(您同意或不同意多少?)
1
1) 安全(您同意或不同意多少?)
Answer 1
10%
Answer 2
22%
Answer 2
19%
Answer 3
15%
Answer 4
8%
Answer 5
14%
Answer 6
14%
2) 控制(您同意或不同意多少?)
Answer 7
22%
Answer 8
23%
Answer 8
18%
Answer 9
13%
Answer 10
5%
Answer 11
9%
Answer 12
10%

在图表上显示692结果
分享, %
40
35
30
25
20
15
10
5
0
在商业自己的农场家族企业启动个人业务业务流程管理项目管理社交网络互利双赢的行为合成
管理风格 手动 常规的 创新的
在生活中生存家庭关系统治力真理的力量竞赛人际关系灵活的流全球视野
颜色米色紫色红色蓝色橙色绿色黄色青绿色
按投票平均, %0441320172318
國家
好的
取消
0561422172214
語言
好的
取消
0551320172317
5 10 15 20 25 30 35 40
圖表
选择要导出的结果(电子邮件)
选择要导出的结果(电子邮件)
所有
vuca-1@sdtest.org.cn
vuca-2@sdtest.org.cn
vuca-3@sdtest.org.cn
vuca-4@sdtest.org.cn
vuca-5@sdtest.org.cn
vuca-6@sdtest.org.cn
vuca-7@sdtest.org.cn
vuca-8@sdtest.org.cn
vuca-9@sdtest.org.cn
vuca-10@sdtest.org.cn
vuca-11@sdtest.org.cn
vuca-12@sdtest.org.cn
vuca-13@sdtest.org.cn
vuca-14@sdtest.org.cn
vuca-15@sdtest.org.cn
vuca-16@sdtest.org.cn
vuca-17@sdtest.org.cn
vuca-18@sdtest.org.cn
vuca-19@sdtest.org.cn
vuca-20@sdtest.org.cn
vuca-21@sdtest.org.cn
vuca-22@sdtest.org.cn
vuca-23@sdtest.org.cn
vuca-24@sdtest.org.cn
vuca-25@sdtest.org.cn
vuca-26@sdtest.org.cn
vuca-27@sdtest.org.cn
vuca-28@sdtest.org.cn
vuca-29@sdtest.org.cn
vuca-30@sdtest.org.cn
vuca-31@sdtest.org.cn
vuca-32@sdtest.org.cn
vuca-33@sdtest.org.cn
vuca-34@sdtest.org.cn
vuca-35@sdtest.org.cn
vuca-36@sdtest.org.cn
vuca-37@sdtest.org.cn
vuca-38@sdtest.org.cn
vuca-39@sdtest.org.cn
vuca-40@sdtest.org.cn
vuca-41@sdtest.org.cn
vuca-42@sdtest.org.cn
vuca-43@sdtest.org.cn
vuca-44@sdtest.org.cn
vuca-45@sdtest.org.cn
vuca-46@sdtest.org.cn
vuca-47@sdtest.org.cn
vuca-48@sdtest.org.cn
vuca-49@sdtest.org.cn
vuca-50@sdtest.org.cn
vuca-51@sdtest.org.cn
vuca-52@sdtest.org.cn
vuca-53@sdtest.org.cn
vuca-54@sdtest.org.cn
vuca-55@sdtest.org.cn
vuca-56@sdtest.org.cn
vuca-57@sdtest.org.cn
vuca-58@sdtest.org.cn
vuca-59@sdtest.org.cn
vuca-60@sdtest.org.cn
vuca-61@sdtest.org.cn
vuca-62@sdtest.org.cn
vuca-63@sdtest.org.cn
vuca-64@sdtest.org.cn
vuca-65@sdtest.org.cn
vuca-66@sdtest.org.cn
vuca-67@sdtest.org.cn
vuca-68@sdtest.org.cn
vuca-69@sdtest.org.cn
vuca-70@sdtest.org.cn
vuca-71@sdtest.org.cn
vuca-72@sdtest.org.cn
vuca-73@sdtest.org.cn
vuca-74@sdtest.org.cn
vuca-75@sdtest.org.cn
vuca-76@sdtest.org.cn
vuca-77@sdtest.org.cn
vuca-78@sdtest.org.cn
vuca-79@sdtest.org.cn
vuca-80@sdtest.org.cn
vuca-81@sdtest.org.cn
vuca-82@sdtest.org.cn
vuca-83@sdtest.org.cn
vuca-84@sdtest.org.cn
vuca-85@sdtest.org.cn
vuca-86@sdtest.org.cn
vuca-87@sdtest.org.cn
vuca-88@sdtest.org.cn
vuca-89@sdtest.org.cn
vuca-90@sdtest.org.cn
vuca-91@sdtest.org.cn
vuca-92@sdtest.org.cn
vuca-93@sdtest.org.cn
vuca-94@sdtest.org.cn
vuca-95@sdtest.org.cn
vuca-96@sdtest.org.cn
vuca-97@sdtest.org.cn
vuca-98@sdtest.org.cn
vuca-99@sdtest.org.cn
vuca-100@sdtest.org.cn
vuca-101@sdtest.org.cn
vuca-102@sdtest.org.cn
vuca-103@sdtest.org.cn
vuca-104@sdtest.org.cn
vuca-105@sdtest.org.cn
vuca-106@sdtest.org.cn
vuca-107@sdtest.org.cn
vuca-108@sdtest.org.cn
vuca-109@sdtest.org.cn
vuca-110@sdtest.org.cn
vuca-111@sdtest.org.cn
vuca-112@sdtest.org.cn
vuca-113@sdtest.org.cn
vuca-114@sdtest.org.cn
vuca-115@sdtest.org.cn
vuca-116@sdtest.org.cn
vuca-117@sdtest.org.cn
vuca-118@sdtest.org.cn
vuca-119@sdtest.org.cn
vuca-120@sdtest.org.cn
vuca-121@sdtest.org.cn
vuca-122@sdtest.org.cn
vuca-123@sdtest.org.cn
vuca-124@sdtest.org.cn
vuca-125@sdtest.org.cn
vuca-126@sdtest.org.cn
vuca-127@sdtest.org.cn
vuca-128@sdtest.org.cn
vuca-129@sdtest.org.cn
vuca-130@sdtest.org.cn
vuca-131@sdtest.org.cn
vuca-132@sdtest.org.cn
vuca-133@sdtest.org.cn
vuca-134@sdtest.org.cn
vuca-135@sdtest.org.cn
vuca-136@sdtest.org.cn
vuca-137@sdtest.org.cn
vuca-138@sdtest.org.cn
vuca-139@sdtest.org.cn
vuca-140@sdtest.org.cn
vuca-141@sdtest.org.cn
vuca-142@sdtest.org.cn
vuca-143@sdtest.org.cn
vuca-144@sdtest.org.cn
vuca-145@sdtest.org.cn
vuca-146@sdtest.org.cn
vuca-147@sdtest.org.cn
vuca-148@sdtest.org.cn
vuca-149@sdtest.org.cn
vuca-150@sdtest.org.cn
vuca-151@sdtest.org.cn
vuca-152@sdtest.org.cn
vuca-153@sdtest.org.cn
vuca-154@sdtest.org.cn
vuca-155@sdtest.org.cn
vuca-156@sdtest.org.cn
vuca-157@sdtest.org.cn
vuca-158@sdtest.org.cn
vuca-159@sdtest.org.cn
vuca-160@sdtest.org.cn
vuca-161@sdtest.org.cn
vuca-162@sdtest.org.cn
vuca-163@sdtest.org.cn
vuca-164@sdtest.org.cn
vuca-165@sdtest.org.cn
vuca-166@sdtest.org.cn
vuca-167@sdtest.org.cn
vuca-168@sdtest.org.cn
vuca-169@sdtest.org.cn
vuca-170@sdtest.org.cn
vuca-171@sdtest.org.cn
vuca-172@sdtest.org.cn
vuca-173@sdtest.org.cn
vuca-174@sdtest.org.cn
vuca-175@sdtest.org.cn
vuca-176@sdtest.org.cn
vuca-177@sdtest.org.cn
vuca-178@sdtest.org.cn
vuca-179@sdtest.org.cn
vuca-180@sdtest.org.cn
vuca-181@sdtest.org.cn
vuca-182@sdtest.org.cn
vuca-183@sdtest.org.cn
vuca-184@sdtest.org.cn
vuca-185@sdtest.org.cn
vuca-186@sdtest.org.cn
vuca-187@sdtest.org.cn
vuca-188@sdtest.org.cn
vuca-189@sdtest.org.cn
vuca-190@sdtest.org.cn
vuca-191@sdtest.org.cn
vuca-192@sdtest.org.cn
vuca-193@sdtest.org.cn
vuca-194@sdtest.org.cn
vuca-195@sdtest.org.cn
vuca-196@sdtest.org.cn
vuca-197@sdtest.org.cn
vuca-198@sdtest.org.cn
vuca-199@sdtest.org.cn
vuca-200@sdtest.org.cn
vuca-201@sdtest.org.cn
vuca-202@sdtest.org.cn
vuca-203@sdtest.org.cn
vuca-204@sdtest.org.cn
vuca-205@sdtest.org.cn
vuca-206@sdtest.org.cn
vuca-207@sdtest.org.cn
vuca-208@sdtest.org.cn
vuca-209@sdtest.org.cn
vuca-210@sdtest.org.cn
vuca-211@sdtest.org.cn
vuca-212@sdtest.org.cn
vuca-213@sdtest.org.cn
vuca-214@sdtest.org.cn
vuca-215@sdtest.org.cn
vuca-216@sdtest.org.cn
vuca-217@sdtest.org.cn
vuca-218@sdtest.org.cn
vuca-219@sdtest.org.cn
vuca-220@sdtest.org.cn
vuca-221@sdtest.org.cn
vuca-222@sdtest.org.cn
vuca-223@sdtest.org.cn
vuca-224@sdtest.org.cn
vuca-225@sdtest.org.cn
vuca-226@sdtest.org.cn
vuca-227@sdtest.org.cn
vuca-228@sdtest.org.cn
vuca-229@sdtest.org.cn
vuca-230@sdtest.org.cn
vuca-231@sdtest.org.cn
vuca-232@sdtest.org.cn
vuca-233@sdtest.org.cn
vuca-234@sdtest.org.cn
vuca-235@sdtest.org.cn
vuca-236@sdtest.org.cn
vuca-237@sdtest.org.cn
vuca-238@sdtest.org.cn
vuca-239@sdtest.org.cn
vuca-240@sdtest.org.cn
vuca-241@sdtest.org.cn
vuca-242@sdtest.org.cn
vuca-243@sdtest.org.cn
vuca-244@sdtest.org.cn
vuca-245@sdtest.org.cn
vuca-246@sdtest.org.cn
vuca-247@sdtest.org.cn
vuca-248@sdtest.org.cn
vuca-249@sdtest.org.cn
vuca-250@sdtest.org.cn
vuca-251@sdtest.org.cn
vuca-252@sdtest.org.cn