AI and the Resilience of Physical Engineering: Why the Future Looks Bright for Men of Understanding
The fundamental mistake in predicting the future of work is assuming that whatever AI is capable of doing, humans will necessarily allow AI to do. Capability is not the same thing as authority. Would society hand an AI independent launch authority over a nuclear arsenal? Would we permit an AI to make the final determination of whether a patient lives or dies? In high-stakes domains, the answer depends not only on what a machine can do, but on what society is willing to delegate. As AI becomes increasingly capable of performing intellectual and technical tasks, education can become more valuable—not because educated people must compete with AI at performing every task themselves, but because society will continue to need people capable of understanding, supervising, validating, and taking responsibility for increasingly powerful systems. Roles involving scientific judgment, legal accountability, institutional authority, ethical responsibility, and consequential decision-making are therefore less likely to disappear simply because AI can perform many of their underlying tasks. Automation can reduce the amount of human labor required to produce an outcome without eliminating the need for humans who possess the knowledge and authority to determine whether that outcome should be trusted, accepted, or acted upon.
I. As the Supply of Digitized Knowledge Rises, Demand for Physical Engineering Increases
The same principle applies to engineering. As artificial intelligence automates an increasing share of software, programming, and other routine digital knowledge work, demand for engineers is likely to increase rather than disappear. AI can make computation and software development dramatically cheaper, but it does not remove the physical constraints imposed by electricity, materials, thermodynamics, manufacturing, structures, aircraft, machines, and infrastructure. Instead, as digital intelligence becomes more abundant, the limiting factor increasingly shifts toward the ability to turn that intelligence into reliable physical systems. Electrical engineers are needed to design and expand the power, electronics, and semiconductor systems on which AI depends; mechanical engineers are needed to design the machines, manufacturing systems, and physical hardware that embody increasingly capable technology; and aerospace engineers are needed to turn advanced computation into aircraft and spacecraft that must operate within unforgiving physical constraints. The lesson of automation is therefore not that higher education has become obsolete or that people should simply abandon intellectual study in favor of manual work. It is that the nature of valuable expertise is changing. An economy with abundant machine intelligence still needs people with the mathematics, physics, engineering knowledge, and systems-level judgment required to design, build, test, and ultimately take responsibility for the physical technologies that intelligence makes possible. Recent federal projections illustrate this shift in demand:
The overall demand for electricity is expected to increase over the projections period due in large part to AI adoption and demand for related infrastructure, such as data centers (U.S. Bureau of Labor Statistics [BLS], 2026).
Electricity consumption reaches record levels in our forecast, driven by data center development and increased manufacturing activity in the commercial and industrial sectors (U.S. Energy Information Administration [EIA], 2026).
The Needs Study report shows that there is a pressing need for additional electric transmission infrastructure due to load growth from data centers, expanding domestic manufacturing, large industrial loads, and a growing economy (U.S. Department of Energy [DOE], 2026).
This exponential growth places strain on grid infrastructure, which can hinder timely construction of desired computing capacity (Granderson et al., 2026).
II. Careers Where Knowledge Meets Physical Reality
Some of the clearest examples of this principle lie at the frontier of science and engineering, where knowledge must be tested against reality rather than merely generated on a screen. Laboratory scientists do not simply reason about the physical world; they interact with it to discover things that were not previously known. They may have to build or modify experimental equipment, prepare materials, operate instruments, control temperatures and pressures, collect measurements, troubleshoot failures, and determine whether an unexpected result reflects a genuine phenomenon or an experimental error. Skilled trades require substantial physical skill, judgment, and technical knowledge, and an automated economy will continue to need people capable of constructing, maintaining, repairing, and operating physical systems. But frontier science and engineering perform an additional function: they create new knowledge and new technological capabilities. The distinction matters because AI cannot simply infer an empirical result that has never been measured or guarantee that a design will behave correctly under physical conditions it has never encountered. Electrical, mechanical, aerospace, and other engineering disciplines therefore occupy an important position between abstract computation and physical reality: they take scientific knowledge and computational tools and turn them into systems that must actually work in the world, under real constraints, with real consequences for failure:
| Career family | Representative BLS occupation | 2025 employment | Median pay | 90th-percentile pay | Annual openings | Projected growth | New jobs | Median worker age | 55+ share | Federal pay | Typical entry education |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Clinical medicine | Physicians & surgeons | 862,800 | $275,930 | >$488,320 | 22,100 | 4% | 31,300 | — | — | — | Doctoral/professional |
| Electrical / electronics engineering | Electrical & electronics engineers | 297,900 | $125,040† | >$184,300 / >$206,960† | 16,300 | 8% | 23,400 | 41.4 | ~24% | $135,890‡ | Bachelor’s |
| Computer hardware / semiconductor engineering | Computer hardware engineers | 76,100 | $161,740 | >$225,330 | 4,100 | 9% | 6,900 | 42.2 | ~17% | $136,990 | Bachelor’s |
| Aerospace engineering | Aerospace engineers | 68,700 | $134,960 | >$205,890 | 3,800 | 8% | 5,700 | 39.5 | ~17% | $142,630‡ | Bachelor’s |
| Mechanical engineering | Mechanical engineers | 298,500 | $104,110 | >$164,340 | 17,800 | 11% | 33,500 | 41.2 | ~23% | — | Bachelor’s |
| Materials / semiconductor engineering | Materials engineers | 23,800 | $112,860 | >$175,720 | 1,300 | 8% | 1,800 | 40.8 | ~23% | — | Bachelor’s |
| Civil / structural engineering | Civil engineers | 380,600 | $100,840 | >$163,220 | 22,700 | 6% | 24,500 | 41.3 | ~21% | $117,790‡ | Bachelor’s |
| Health / safety engineering | Health and safety engineers | 22,300 | $115,160 | >$171,250 | 1,200 | 5% | 1,200 | — | ~22%§ | — | Bachelor’s |
| Biomedical engineering | Bioengineers & biomedical engineers | 23,800 | $109,370 | >$168,180 | 1,200 | 8% | 1,800 | — | ~21% | — | Bachelor’s / graduate for some roles |
| Nuclear engineering / nuclear technology | Nuclear engineers | 16,600 | $133,970 | >$196,290 | 800 | 0% | 100 | — | — | $118,440‡ | Bachelor’s |
| Chemical / process engineering | Chemical engineers | 21,900 | $125,040 | >$182,880 | 1,100 | 5% | 1,000 | 38.1 | ~22% | $133,050‡ | Bachelor’s |
| Experimental chemistry / materials science | Chemists & materials scientists | 93,500 | $95,060¶ | >$160,830 / >$197,290¶ | 6,500 | 7% | 6,200 | 39.8 | ~22% | $137,720‡¶ | Bachelor’s / graduate for research |
| Biophysics / experimental life science | Biochemists & biophysicists | 35,200 | $127,410 | >$201,110 | 2,900 | 12% | 4,300 | — | — | — | Doctoral/professional |
| Medical / biomedical R&D science | Medical scientists | 181,000 | $103,410 | >$177,780 | 10,500 | 13% | 22,700 | 41.2 | ~23% | — | Doctoral/professional |
| Physics / physical-science R&D | Physicists | 23,200 | $172,250 | >$274,110 | ~1,500¶ | 7% | 1,700 | — | — | $146,900‡ | PhD research / bachelor’s federal |
| Geoscience / field physical science | Geoscientists | 25,400 | $101,920 | >$200,230 | 1,800 | 5% | 1,300 | — | ~23% | $119,430‡ | Bachelor’s |
| Atmospheric / observational science | Atmospheric scientists, including meteorologists | 10,700 | $99,070 | >$161,890 | 800 | 3% | 300 | — | ~27% | $122,700‡ | Bachelor’s / graduate for research |
‡ Industry/federal figures are occupation-specific median wages, not the occupation-wide median. Examples include $142,630 for aerospace engineers, $117,790 for civil engineers, $119,430 for geoscientists, $122,700 for atmospheric scientists, $124,870 for computer and information research scientists, and $146,900 for physicists in the federal government. Where BLS does not publish a directly comparable federal-government figure on the occupation page, the table uses — rather than substituting a broader government-industry figure.
§ BLS's Current Population Survey age table does not publish a separate age series for health and safety engineers; the approximate 22% 55+ figure is based on the broader industrial engineers, including health and safety CPS category and should therefore be treated as a proxy rather than an exact health-and-safety-engineer estimate.
¶ The chemists/materials-scientists family combines two BLS occupations. BLS reports a combined $95,060 median for the family, but the highest-10-percent thresholds are published separately: >$160,830 for chemists and >$197,290 for materials scientists. The federal figure shown, $137,720, is specifically the federal-government median for chemists; BLS does not provide an equivalent federal figure for materials scientists on the same profile.
III. Labor Market Outcomes of College Graduates by Major
The New York Fed's 2024 labor-market outcomes by major are striking. Nearly two years after the public release of ChatGPT, the majors most closely associated with advanced technical work remained among those with the lowest underemployment rates in the country. Aerospace engineering stood at 14.7%, civil engineering at 15.6%, computer engineering at 15.8%, chemical engineering at 17.9%, computer science at 19.1%, mechanical engineering at 20.1%, and electrical engineering at 21.1%, against 39.4% across all majors. Their early- and mid-career wages were likewise substantially above the college-graduate average. Whatever expectations there may have been that generative AI would rapidly undermine the value of technically demanding education, these 2024 outcomes show no such collapse:
| Major | Unemployment Rate | Underemployment Rate | Median Wage Early Career | Median Wage Mid-Career | Share with Graduate Degree |
|---|---|---|---|---|---|
| Nursing | 2.1% | 12.8% | $70,000 | $87,000 | 30.3% |
| Aerospace Engineering | 2.2% | 14.7% | $85,000 | $130,000 | 45.5% |
| Civil Engineering | 2.3% | 15.6% | $75,000 | $115,000 | 37.1% |
| Computer Engineering | 7.8% | 15.8% | $90,000 | $131,000 | 39.4% |
| Special Education | 0.7% | 16.0% | $46,000 | $56,000 | 61.1% |
| Elementary Education | 1.2% | 16.2% | $45,000 | $55,000 | 50.6% |
| Chemical Engineering | 4.7% | 17.9% | $85,000 | $135,000 | 48.1% |
| Construction Services | 2.2% | 17.9% | $75,000 | $120,000 | 12.2% |
| Computer Science | 7.0% | 19.1% | $87,000 | $120,000 | 32.7% |
| Mechanical Engineering | 4.4% | 20.1% | $80,000 | $120,000 | 39.1% |
| Electrical Engineering | 3.2% | 21.1% | $82,000 | $123,000 | 47.7% |
| Accounting | 2.6% | 21.2% | $68,000 | $97,000 | 33.7% |
| Secondary Education | 2.1% | 21.7% | $45,300 | $62,000 | 52.7% |
| Early Childhood Education | 6.6% | 22.3% | $45,000 | $52,000 | 43.8% |
| Miscellaneous Physical Sciences | 2.5% | 22.5% | $60,000 | $81,000 | 54.0% |
| Miscellaneous Education | 1.1% | 24.5% | $48,000 | $60,000 | 59.9% |
| Information Systems & Management | 6.0% | 25.6% | $67,000 | $100,000 | 27.9% |
| Architecture | 6.8% | 25.7% | $60,000 | $91,000 | 40.8% |
| Mathematics | 5.8% | 26.2% | $70,000 | $100,000 | 51.3% |
| Miscellaneous Engineering | 3.7% | 26.4% | $75,000 | $105,000 | 46.3% |
| Finance | 2.8% | 27.8% | $70,000 | $112,000 | 31.5% |
| General Education | 2.6% | 28.1% | $45,000 | $56,000 | 51.7% |
| Business Analytics | 5.0% | 28.7% | $72,000 | $109,000 | 26.7% |
| Physics | 6.6% | 29.1% | $67,000 | $105,000 | 67.3% |
| Social Services | 1.9% | 30.1% | $43,000 | $60,000 | 53.8% |
| Pharmacy | 5.6% | 30.5% | $40,000 | $85,000 | 67.4% |
| General Engineering | 4.5% | 31.1% | $75,000 | $105,000 | 37.1% |
| Earth Sciences | 3.8% | 31.4% | $50,000 | $75,000 | 45.6% |
| Industrial Engineering | 4.2% | 31.7% | $83,000 | $100,000 | 40.6% |
| Economics | 3.5% | 33.1% | $72,000 | $115,000 | 41.6% |
| Commercial Art & Graphic Design | 5.7% | 37.1% | $50,000 | $80,000 | 13.0% |
| General Social Sciences | 2.3% | 38.0% | $45,200 | $75,000 | 39.3% |
| Advertising and Public Relations | 5.7% | 38.1% | $60,000 | $92,000 | 22.8% |
| Treatment Therapy | 2.6% | 39.1% | $50,000 | $70,000 | 53.4% |
| Biochemistry | 2.7% | 42.0% | $52,000 | $100,000 | 71.9% |
| Chemistry | 4.3% | 42.8% | $50,000 | $86,000 | 67.5% |
| Geography | 1.6% | 43.0% | $50,000 | $88,000 | 34.7% |
| Journalism | 2.3% | 43.3% | $49,000 | $87,000 | 26.3% |
| Engineering Technologies | 1.7% | 44.4% | $65,000 | $104,000 | 27.9% |
| Health Services | 3.8% | 45.1% | $50,000 | $67,000 | 52.9% |
| International Affairs | 6.1% | 45.2% | $64,000 | $90,000 | 46.3% |
| Interdisciplinary Studies | 3.9% | 45.2% | $52,000 | $80,000 | 40.4% |
| Art History | 6.7% | 45.3% | $45,000 | $91,000 | 47.8% |
| Public Policy and Law | 2.2% | 45.5% | $55,000 | $80,000 | 47.0% |
| Family and Consumer Sciences | 4.3% | 45.6% | $48,000 | $65,000 | 35.7% |
| Miscellaneous Biological Science | 4.4% | 46.9% | $46,000 | $80,000 | 59.5% |
| Medical Technicians | 6.2% | 47.0% | $62,000 | $80,000 | 24.7% |
| Philosophy | 5.1% | 47.1% | $52,000 | $80,000 | 56.9% |
| Theology and Religion | 3.1% | 47.5% | $41,600 | $66,000 | 46.4% |
| Psychology | 5.0% | 48.3% | $45,000 | $72,000 | 51.9% |
| Nutrition Sciences | 4.5% | 48.3% | $50,000 | $70,000 | 50.6% |
| English Language | 6.1% | 48.5% | $48,000 | $76,000 | 46.2% |
| Political Science | 4.5% | 48.7% | $52,000 | $100,000 | 52.7% |
| Miscellaneous Technologies | 3.6% | 49.0% | $54,000 | $96,000 | 19.5% |
| Marketing | 4.4% | 49.3% | $60,000 | $100,000 | 19.7% |
| Ethnic Studies | 4.9% | 49.6% | $52,000 | $90,000 | 51.2% |
| History | 4.3% | 50.1% | $47,500 | $80,000 | 51.4% |
| General Business | 4.3% | 50.4% | $60,000 | $90,000 | 25.3% |
| Environmental Studies | 6.3% | 50.5% | $50,000 | $80,000 | 31.9% |
| Biology | 4.3% | 51.1% | $45,000 | $83,000 | 64.0% |
| Sociology | 4.6% | 52.0% | $49,900 | $79,000 | 39.1% |
| Mass Media | 5.2% | 52.1% | $50,000 | $80,000 | 19.3% |
| Business Management | 3.8% | 52.6% | $56,000 | $82,000 | 25.3% |
| Communications | 3.9% | 53.0% | $52,000 | $88,000 | 24.7% |
| Animal and Plant Sciences | 2.5% | 53.5% | $46,600 | $79,000 | 36.6% |
| Foreign Language | 1.6% | 54.0% | $48,000 | $77,000 | 53.1% |
| Liberal Arts | 3.8% | 54.6% | $45,000 | $75,000 | 30.9% |
| Anthropology | 7.9% | 55.3% | $45,000 | $65,000 | 48.2% |
| Agriculture | 1.4% | 57.1% | $56,000 | $80,000 | 22.5% |
| Leisure and Hospitality | 2.7% | 58.1% | $45,000 | $75,000 | 36.1% |
| Fine Arts | 7.7% | 58.9% | $45,000 | $72,000 | 22.5% |
| Performing Arts | 7.0% | 63.9% | $44,000 | $75,000 | 38.4% |
| Criminal Justice | 3.6% | 65.8% | $50,000 | $80,000 | 24.6% |
| Overall | 4.2% | 39.4% | $58,000 | $87,000 | 39.3% |
IV. Labor Market Outcomes for AI-Exposed College Majors
A September 2026 working paper from the U.S. Census Bureau’s Center for Economic Studies finds that the effects of generative AI are already diverging across college majors. Graduates from the most AI-exposed fields experienced significant declines in early-career employment and earnings following the introduction of ChatGPT, with computer science, computer engineering, and computer programming among the fields experiencing the largest negative outcomes. At the same time, several engineering disciplines—including civil, petroleum, environmental, and energy systems engineering—showed positive employment outcomes. Mechanical engineering appears comparatively resilient, with relatively low AI exposure, only a modest employment decline, and positive earnings, while electrical engineering has substantially higher AI exposure but only a similarly modest employment decline and slightly positive earnings. Together, these results suggest that some engineering fields remain comparatively resilient even as AI increasingly automates parts of engineering work (Orr et al., 2026, pp. 56-59). The following STEM disciplines are sorted by their 2024 employment coefficients in Table A3:
| Rank | Field of degree | N2018–22 | Beta | Core | Usage | Share top | Employment (SE) | Earnings (SE) |
|---|---|---|---|---|---|---|---|---|
| 1 | General Engineering | 2,600 | 0.40 | 0.40 | -0.77 | 0.175 | +0.050 (0.006) | +0.051 (0.017) |
| 2 | Petroleum Engineering | 3,600 | 0.51 | 0.25 | -1.18 | 0.139 | +0.038 (0.004) | +0.047 (0.007) |
| 3 | Civil Engineering | 25,500 | 0.16 | -0.05 | -1.81 | 0.059 | +0.033 (0.001) | +0.077 (0.003) |
| 4 | Medical Technologies Technicians | 17,000 | -1.41 | -1.41 | -1.82 | 0.045 | +0.030 (0.002) | +0.068 (0.006) |
| 5 | Environmental Engineering | 2,400 | 0.35 | 0.58 | -1.56 | 0.079 | +0.028 (0.002) | +0.022 (0.006) |
| 6 | Engineering Technologies | 5,500 | 0.29 | 0.40 | -0.16 | 0.258 | +0.024 (0.003) | +0.041 (0.006) |
| 7 | Biochemical Sciences | 13,000 | -0.17 | -0.20 | -0.48 | 0.136 | +0.018 (0.001) | -0.044 (0.004) |
| 8 | Industrial Production Technologies | 7,100 | -0.17 | 0.09 | -1.36 | 0.136 | +0.018 (0.003) | +0.026 (0.007) |
| 9 | Zoology | 4,000 | -1.20 | -1.06 | -1.22 | 0.110 | +0.017 (0.003) | +0.124 (0.010) |
| 10 | Animal Sciences | 18,000 | -1.31 | -1.09 | -1.08 | 0.133 | +0.017 (0.002) | +0.068 (0.005) |
| 11 | Energy Systems Engineering | 3,700 | 0.26 | 0.18 | -1.15 | 0.115 | +0.014 (0.002) | +0.063 (0.006) |
| 12 | Neuroscience | 7,500 | -0.04 | -0.15 | -0.44 | 0.160 | +0.009 (0.003) | +0.031 (0.006) |
| 13 | Agriculture Production and Management | 19,000 | -0.33 | -0.44 | -0.44 | 0.217 | +0.009 (0.002) | +0.047 (0.005) |
| 14 | Microbiology and Immunology | 4,600 | -0.28 | -0.30 | -0.52 | 0.155 | +0.005 (0.002) | +0.056 (0.005) |
| 15 | Natural Resources Management | 8,800 | -0.75 | -0.57 | -1.03 | 0.120 | +0.005 (0.002) | +0.063 (0.006) |
| 16 | Biology | 129,000 | -0.48 | -0.51 | -0.42 | 0.164 | +0.004 (0.001) | +0.001 (0.002) |
| 17 | Multi-disciplinary or General Science | 70,000 | -0.52 | -0.59 | -0.41 | 0.170 | +0.003 (0.002) | -0.009 (0.005) |
| 18 | Molecular Biology | 4,900 | -0.10 | -0.14 | -0.60 | 0.137 | +0.003 (0.004) | -0.065 (0.009) |
| 19 | Chemical Engineering | 18,500 | 0.61 | 0.91 | -1.14 | 0.120 | 0.000 (0.002) | +0.039 (0.005) |
| 20 | Plant Science and Agronomy | 4,100 | -1.23 | -1.08 | -1.28 | 0.094 | -0.003 (0.002) | -0.025 (0.005) |
| 21 | Atmospheric Sciences and Meteorology | 1,400 | 0.48 | 0.54 | -1.01 | 0.146 | -0.004 (0.003) | -0.143 (0.006) |
| 22 | Genetics | 1,900 | -0.37 | -0.51 | -0.83 | 0.094 | -0.004 (0.004) | +0.020 (0.007) |
| 23 | Environmental Science | 14,500 | -0.30 | -0.09 | -0.58 | 0.156 | -0.004 (0.001) | +0.005 (0.003) |
| 24 | Architectural Engineering | 1,200 | 0.18 | 0.13 | -1.39 | 0.119 | -0.006 (0.004) | +0.089 (0.009) |
| 25 | Electrical Engineering Technology | 2,700 | 0.54 | 0.45 | -0.25 | 0.236 | -0.006 (0.003) | +0.117 (0.006) |
| 26 | Energy Systems Technologies/Technicians | 15,000 | 0.20 | 0.14 | -0.87 | 0.207 | -0.008 (0.002) | +0.067 (0.004) |
| 27 | Mechanical Engineering | 59,000 | 0.53 | 0.59 | -1.21 | 0.110 | -0.009 (0.001) | +0.039 (0.003) |
| 28 | Mechanical Engineering Related Technologies | 6,100 | -0.11 | 0.11 | -1.51 | 0.122 | -0.010 (0.003) | +0.048 (0.007) |
| 29 | Information Sciences | 18,000 | 2.15 | 2.27 | 1.61 | 0.523 | -0.011 (0.002) | -0.079 (0.007) |
| 30 | Electrical Engineering | 27,000 | 1.48 | 1.10 | -0.39 | 0.267 | -0.011 (0.001) | +0.016 (0.003) |
| 31 | Applied Mathematics | 2,700 | 1.74 | 1.57 | 1.33 | 0.383 | -0.011 (0.003) | -0.101 (0.008) |
| 32 | Engineering Mechanics, Physics, and Science | 1,700 | 0.92 | 1.06 | -0.44 | 0.230 | -0.012 (0.004) | -0.008 (0.009) |
| 33 | Materials Engineering | 4,500 | 0.42 | 0.80 | -0.98 | 0.139 | -0.014 (0.003) | +0.026 (0.008) |
| 34 | Chemistry | 20,500 | -0.20 | -0.12 | -0.49 | 0.122 | -0.014 (0.001) | +0.030 (0.003) |
| 35 | Mathematics | 22,500 | 1.09 | 0.98 | 0.71 | 0.327 | -0.015 (0.001) | -0.051 (0.003) |
| 36 | Industrial and Manufacturing Engineering | 13,500 | 0.80 | 1.15 | -0.40 | 0.195 | -0.019 (0.002) | +0.021 (0.005) |
| 37 | Management Information Systems and Statistics | 27,500 | 1.90 | 2.27 | 1.51 | 0.483 | -0.021 (0.001) | -0.089 (0.003) |
| 38 | Computer Programming | 6,400 | 2.60 | 2.11 | 1.91 | 0.543 | -0.023 (0.003) | -0.113 (0.008) |
| 39 | Miscellaneous Biology | 6,100 | -0.41 | -0.44 | -0.74 | 0.134 | -0.024 (0.003) | +0.017 (0.006) |
| 40 | Ecology | 4,200 | -0.88 | -0.78 | -0.99 | 0.096 | -0.025 (0.004) | +0.066 (0.011) |
| 41 | Statistics | 3,900 | 1.61 | 1.60 | 1.41 | 0.393 | -0.026 (0.002) | -0.081 (0.006) |
| 42 | Physics | 77,000 | 0.85 | 0.66 | 0.03 | 0.249 | -0.027 (0.001) | -0.104 (0.004) |
| 43 | Biomedical Engineering | 12,500 | 0.71 | 0.76 | -0.39 | 0.212 | -0.030 (0.002) | -0.083 (0.005) |
| 44 | Aerospace Engineering | 24,500 | 1.08 | 0.59 | -1.02 | 0.174 | -0.032 (0.003) | -0.091 (0.007) |
| 45 | Computer Administration, Management, and Security | 14,000 | 1.88 | 2.34 | 1.57 | 0.545 | -0.033 (0.003) | -0.096 (0.008) |
| 46 | Biological Engineering | 3,600 | 0.57 | 0.71 | -0.79 | 0.173 | -0.038 (0.003) | +0.103 (0.007) |
| 47 | Geology and Earth Science | 8,300 | -0.09 | -0.10 | -0.77 | 0.133 | -0.040 (0.001) | +0.010 (0.004) |
| 48 | Food Science | 2,300 | -0.08 | 0.16 | -0.63 | 0.128 | -0.040 (0.003) | +0.047 (0.008) |
| 49 | Nuclear Engineering | 1,400 | 0.65 | 0.47 | -1.10 | 0.131 | -0.052 (0.003) | +0.102 (0.009) |
| 50 | Computer and Information Systems—General | 63,000 | 2.07 | 2.19 | 1.39 | 0.556 | -0.054 (0.002) | -0.142 (0.003) |
| 51 | Computer Science | 43,500 | 3.40 | 2.90 | 1.65 | 0.662 | -0.061 (0.001) | -0.190 (0.003) |
| 52 | Computer Engineering | 18,500 | 3.17 | 2.56 | 1.18 | 0.598 | -0.068 (0.002) | -0.120 (0.004) |
V. What Happened After the Initial Shock? Evidence Through 2026
The Census evidence shows that employment outcomes had already begun diverging across fields by 2024, but newer evidence through 2026 suggests that the pattern is becoming clearer without amounting to economy-wide job destruction. Stanford's August 2026 analysis of payroll data covering millions of U.S. workers through June 2026 finds “no evidence of widespread, economy-wide job displacement,” while employment of young workers in highly AI-exposed occupations stands “19% below where it would be had it kept pace with that of their less-exposed peers” (Brynjolfsson et al., 2026). Stanford further finds that the divergence operates primarily through reduced hiring of young workers and is concentrated in occupations where AI use “primarily substitutes for human tasks,” while employment is flat or rising where AI primarily complements workers (Brynjolfsson et al., 2026). Revelio Labs' 2026 labor-market tracker independently finds “continued weakness in junior high-exposure roles” and reports that “87% of how work is changing happens inside jobs, instead of a change in the job mix” (Revelio Labs, 2026). The emerging picture therefore is not that AI simply eliminates occupations, but that it is changing the economic value and composition of particular forms of work, with the strongest pressure appearing where tasks can be reproduced by software and less evidence of wholesale displacement where work remains tied to physical systems, human interaction, or other constraints that software alone cannot remove. This is broadly consistent with the Census results above: the most exposed computing fields experienced some of the weakest outcomes, while several physical engineering fields continued to show positive employment coefficients.
VI. What Does the Next Decade Look Like? The BLS Outlook
The forward-looking evidence points in the same general direction, although with important qualifications. In its new AI-exposure framework accompanying the 2025–2035 employment projections, the Bureau of Labor Statistics combines theoretical measures of AI capability with observed AI-use measures and classifies occupations into four relative exposure categories: Low, Moderate, High, and Very High. BLS explicitly cautions that “Exposure does not imply job loss, productivity gains, automation probability, or wage effects” and that “An exposure category is not a forecast of employment growth or decline” (U.S. Bureau of Labor Statistics [BLS], 2026b). The accompanying employment projections nevertheless provide a useful test of the physical-bottleneck thesis: total U.S. employment is projected to grow 3.5% from 2025 to 2035, while “Overall employment of electrical and electronics engineers is projected to grow 8 percent from 2025 to 2035” (BLS, 2026k), computer hardware engineers are projected to grow 9% (BLS, 2026j), and “Employment of mechanical engineers is projected to grow 11 percent from 2025 to 2035” (BLS, 2026p). This does not mean engineering is immune to AI—indeed, AI is likely to increase engineer productivity and transform the content of engineering work—but it does suggest that making computation and digital intelligence cheaper does not automatically remove the need for the hardware, infrastructure, power systems, manufacturing, machines, testing, certification, and physical deployment that turn computation into useful reality. The relevant economic question is therefore not simply whether AI can perform more engineering tasks, but whether the productivity gains from AI will reduce engineering demand faster than the expansion and increasing complexity of the physical economy create new demand; the employment trends and projections examined above remain consistent with the possibility that AI’s cheapening of digital intelligence will increase, rather than diminish, the relative importance of the physical systems that engineers design, build, test, and maintain.
VII. Predictions for the Future of Work
7.1 Elon Musk — Extreme
The most dramatic forecasts of artificial intelligence go considerably further, predicting that AI and robotics will eventually eliminate the economic necessity of human labor altogether. Elon Musk, for example, has predicted that within roughly ten to twenty years, work could become optional. Such a future is conceivable, but it is fundamentally different from claiming that AI will soon eliminate the economic value of education or engineering. Between today's economy and a hypothetical world in which machines can autonomously reproduce, manufacture, maintain, and improve every physical system on which civilization depends lies an enormous technological and industrial transition. That transition requires electrical power, semiconductors, factories, machines, vehicles, infrastructure, and the engineers capable of designing and integrating them. The possibility that human labor could eventually become unnecessary therefore does not imply that engineers are becoming unnecessary now; indeed, the attempt to automate the physical world may initially increase demand for the expertise required to build it:
Musk: In the long term where will things end up? I don’t know what long term is. Maybe it’s 10-20 years, something like that. For me that’s long term. My prediction is that work will be optional. ... Between now and then there’s actually a lot of work to get to that point. ... My guess is if you go out long enough, assuming there’s a continued improvement in AI and robotics, which there seems likely, the money will stop being relevant at some point in the future. Now there will still be constraints on power like electricity and mass. The fundamental physics elements will still be constraints. But I think at some point currency becomes irrelevant (Shimabukuro, 2025, para. 3).
More importantly, even this endpoint remains constrained by physical reality. Musk himself acknowledges that “there will still be constraints on power,” including electricity and mass, and that “the fundamental physics elements will still be constraints.” These are not peripheral details; they are engineering problems. An increasingly automated civilization would still have to generate electricity, manufacture machines, process materials, expand infrastructure, and ensure that physical systems operate reliably at scale. AI may eventually automate much of the work engineers perform, but the expansion of AI itself requires people capable of solving the physical and technical constraints that make that expansion possible. The more aggressively civilization automates the physical world, the greater the engineering required to build and sustain it.
7.2 Jensen Huang — Moderate
Jensen Huang offers a more moderate forecast of AI's effect on employment. While he expects artificial intelligence to fundamentally transform the labor market and eliminate or alter some jobs, he rejects the idea that increasing automation will necessarily lead to the widespread disappearance of human work. Instead, Huang argues that AI will make workers more productive, change the nature of existing occupations, and create new forms of economic activity, allowing employment to continue expanding even as individual tasks become automated:
First of all, I think the narratives of AI destroying jobs is not going to help America. Yeah. First of all, it’s just false. Of course, with every technology, and every single day that goes by, jobs of the past are changed (Shimabukuro, 2026, 27:33).
There'll be more people working at the end of this industrial revolution than at the beginning of it. Just like at the end of the last one, the beginning of this one (Shimabukuro, 2026, 32:18).
... it is unlikely most people will lose a job to AI. It is most likely that most people will lose their job to somebody who uses AI (Shimabukuro, 2026, 36:03).
7.3 Jeff Bezos — Optimistic
Jeff Bezos offers an even more optimistic forecast of AI's effect on employment. Rather than viewing AI primarily as a technology that substitutes for human labor, Bezos argues that its productivity gains will elevate workers and expand economic activity. He rejects the idea that AI-driven productivity will necessarily produce a permanent surplus of workers, instead arguing that greater productivity can create new opportunities and demand for human labor. His view is that AI will not simply replace workers, but enable them to accomplish more while increasing the productive capacity of the economy:
These people are wrong. So what's really going to happen is that it's going to elevate all of these people (CNBC, 2026a, paras. 441–442).
We're going to have so much productivity in our economy... (CNBC, 2026a, para. 443).
I think that there's going to be a labor shortage as a result (CNBC, 2026a, para. 435).
No, I know there's a lot of concern in general about AI and job loss. I have a very different view. I think that what's – I think, what's actually going to happen is we're going to have labor scarcity as a result. People are going to have to work hard (CNBC, 2026b, para. 201).
Conclusion
The rise of artificial intelligence is therefore not a reason to abandon higher education or technical specialization. As digital intelligence becomes more abundant, the value of expertise may shift toward human judgment, institutional responsibility, scientific discovery, and the engineering required to turn computation into reliable physical systems. AI may ultimately automate much of the work engineers perform, but an increasingly automated civilization must still be built—and building it remains an enormous engineering problem.
- Brynjolfsson, E., Chandar, B., & Chen, R. (2026, August 12). Canaries in the coal mine? Six facts about the recent employment effects of artificial intelligence. Stanford Digital Economy Lab. https://digitaleconomy.stanford.edu/publication/canaries-in-the-coal-mine-six-facts-about-the-recent-employment-effects-of-artificial-intelligence/
- CNBC. (2026a, May 20). CNBC exclusive: Transcript: Jeff Bezos speaks with CNBC’s Andrew Ross Sorkin on “Squawk Box” today. Versant Media. https://pressroom.versantmedia.com/cnbc/press-releases/cnbc-exclusive-transcript-jeff-bezos-speaks-cnbcs-andrew-ross-sorkin-squawk-box
- CNBC. (2026b, June 11). CNBC exclusive: Transcript: Prometheus co-founders and co-CEOs Jeff Bezos and Vik Bajaj speak with CNBC’s David Faber on “Squawk on the Street” today. Versant Media. https://pressroom.versantmedia.com/cnbc/press-releases/cnbc-exclusive-transcript-prometheus-co-founders-and-co-ceos-jeff-bezos-and-vik
- Federal Reserve Bank of New York. (2026). The labor market for recent college graduates. https://www.newyorkfed.org/research/college-labor-market#--:explore:outcomes-by-major
- Granderson, J., Hoffman, I. M., Holecek, B., Crowe, E., Smith, S. J., & Frick, N. M. (2026). Integrating AI data centers with the power grid. The Bridge. https://doi.org/10.20357/B7X61S
- Orr, C., Tucker, L. C., & Warren, L. (2026). Graduating into disruption: Labor market outcomes for AI-exposed college majors (CES Working Paper No. 26-56). U.S. Census Bureau, Center for Economic Studies. https://www.census.gov/library/working-papers/2026/adrm/CES-WP-26-56.html
- Revelio Labs. (2026, September 3). AI labor market tracker: August 2026. https://www.reveliolabs.com/ai-labor-market-tracker/us/august-2026
- Shimabukuro, J. (2025, November 19). Musk and Huang at US-Saudi Forum 19 Nov 2025: An informal transcript. Educational Technology and Change Journal. https://etcjournal.com/2025/11/19/musk-and-huang-at-us-saudi-forum-19-nov-2025-an-informal-transcript/
- Shimabukuro, J. (2026, May 19). Transcript of Jensen Huang’s “U.S. leadership in AI” talk on 9 April 2026. Educational Technology and Change Journal. https://etcjournal.com/2026/05/19/transcript-of-jensen-huangs-u-s-leadership-in-ai-talk-on-9-april-2026/
- U.S. Bureau of Labor Statistics. (2026a, August 27). Aerospace engineers. Occupational Outlook Handbook. https://www.bls.gov/ooh/architecture-and-engineering/aerospace-engineers.htm
- U.S. Bureau of Labor Statistics. (2026b, August 27). Artificial intelligence (AI) exposure categories. https://www.bls.gov/emp/publications/ai-exposure-categories.htm
- U.S. Bureau of Labor Statistics. (2026c, August 27). Atmospheric scientists, including meteorologists. Occupational Outlook Handbook. https://www.bls.gov/ooh/life-physical-and-social-science/atmospheric-scientists-including-meteorologists.htm
- U.S. Bureau of Labor Statistics. (2026d, August 27). Biochemists and biophysicists. Occupational Outlook Handbook. https://www.bls.gov/ooh/life-physical-and-social-science/Biochemists-and-biophysicists.htm
- U.S. Bureau of Labor Statistics. (2026e, August 27). Bioengineers and biomedical engineers. Occupational Outlook Handbook. https://www.bls.gov/ooh/architecture-and-engineering/biomedical-engineers.htm
- U.S. Bureau of Labor Statistics. (2026f, August 27). Chemical engineers. Occupational Outlook Handbook. https://www.bls.gov/ooh/architecture-and-engineering/chemical-engineers.htm
- U.S. Bureau of Labor Statistics. (2026g, August 27). Chemists and materials scientists. Occupational Outlook Handbook. https://www.bls.gov/ooh/life-physical-and-social-science/chemists-and-materials-scientists.htm
- U.S. Bureau of Labor Statistics. (2026h, August 27). Civil engineers. Occupational Outlook Handbook. https://www.bls.gov/ooh/architecture-and-engineering/civil-engineers.htm
- U.S. Bureau of Labor Statistics. (2026i, August 27). Computer and information research scientists. Occupational Outlook Handbook. https://www.bls.gov/ooh/computer-and-information-technology/computer-and-information-research-scientists.htm
- U.S. Bureau of Labor Statistics. (2026j, August 27). Computer hardware engineers. Occupational Outlook Handbook. https://www.bls.gov/ooh/architecture-and-engineering/computer-hardware-engineers.htm
- U.S. Bureau of Labor Statistics. (2026k, August 27). Electrical and electronics engineers. Occupational Outlook Handbook. https://www.bls.gov/ooh/architecture-and-engineering/electrical-and-electronics-engineers.htm
- U.S. Bureau of Labor Statistics. (2026l). Employed people by detailed occupation and age: 2025 annual averages (Table 11b). Current Population Survey. https://www.bls.gov/cps/cpsaat11b.htm
- U.S. Bureau of Labor Statistics. (2026m, August 27). Geoscientists. Occupational Outlook Handbook. https://www.bls.gov/ooh/life-physical-and-social-science/geoscientists.htm
- U.S. Bureau of Labor Statistics. (2026n, August 27). Health and safety engineers. Occupational Outlook Handbook. https://www.bls.gov/ooh/architecture-and-engineering/health-and-safety-engineers.htm
- U.S. Bureau of Labor Statistics. (2026o, August 27). Materials engineers. Occupational Outlook Handbook. https://www.bls.gov/ooh/architecture-and-engineering/materials-engineers.htm
- U.S. Bureau of Labor Statistics. (2026p, August 27). Mechanical engineers. Occupational Outlook Handbook. https://www.bls.gov/ooh/architecture-and-engineering/mechanical-engineers.htm
- U.S. Bureau of Labor Statistics. (2026q, August 27). Medical scientists. Occupational Outlook Handbook. https://www.bls.gov/ooh/life-physical-and-social-science/medical-scientists.htm
- U.S. Bureau of Labor Statistics. (2026r, August 27). Nuclear engineers. Occupational Outlook Handbook. https://www.bls.gov/ooh/architecture-and-engineering/nuclear-engineers.htm
- U.S. Bureau of Labor Statistics. (2026s, August 27). Physicians and surgeons. Occupational Outlook Handbook. https://www.bls.gov/ooh/healthcare/physicians-and-surgeons.htm
- U.S. Bureau of Labor Statistics. (2026t, August 27). Physicists and astronomers. Occupational Outlook Handbook. https://www.bls.gov/ooh/life-physical-and-social-science/physicists-and-astronomers.htm
- U.S. Department of Energy, Office of Electricity. (2026, July 9). DOE's Office of Electricity publishes 2026 draft National Transmission Needs Study to strengthen America's grid. https://www.energy.gov/oe/articles/does-office-electricity-publishes-2026-draft-national-transmission-needs-study
- U.S. Energy Information Administration. (2026, September 9). Short-term energy outlook. https://www.eia.gov/outlooks/steo/report/elec_coal_renew.php
- U.S. Bureau of Labor Statistics. (2026u, August 27). Employment projections: 2025–2035 summary. U.S. Department of Labor. https://www.bls.gov/news.release/ecopro.nr0.htm
Ordo Luminis Fraternitatis Aeternae