How do BLS benchmark revisions and accelerating AI automation complicate claims of a strong U.S. labour market?

Scott Douglas Jacobsen interviews Rick Rosner on why U.S. job “strength” looks shakier after revisions. Rosner notes the BLS benchmark cut 2025 payroll gains from 584,000 to 181,000; May–December gains totalled 12,000; January 2026 added 130,000—matching a year that “changed little.” He explains why headline messaging eclipses technical revisions, then connects labour resilience to policy: WHO withdrawal and anti-vaccine rhetoric raise outbreak and supply-chain risks; unlawful ICE detentions demand accountability; the SAVE Act could burden eligible voters. He closes on AI: rapid improvements may compress middle-skill work, redefining jobs, careers, and platform income, and forcing unions to renegotiate authorship norms.

Scott Douglas Jacobsen: What is happening with the current employment situation in the United States?

Rick Rosner: The White House says job growth is strong and that America is back. The Bureau of Labour Statistics (BLS) originally reported that the U.S. added 584,000 nonfarm payroll jobs in 2025, but the annual benchmark revision reduced that figure to 181,000.

On a month-to-month basis, the revised BLS data show that from May through December 2025, total payroll gains summed to 12,000 jobs. The January 2026 jobs report shows 130,000 jobs added that month.

Jacobsen: For contrast, what does a healthy U.S. job market typically look like?

Rosner: In a healthy market, payroll growth is often in the range of hundreds of thousands per month, not near-zero for extended periods. BLS data in the January 2026 release characterize 2025 as a year in which payroll employment “changed little,” averaging +15,000 per month.

Jacobsen: Put this in a broader presidential context.

Rosner: Claims about presidents “creating jobs” are complicated because job growth depends on broader economic conditions. Still, net job change over a presidential term provides context. By the end of Trump’s first term, the economy had fewer jobs than at the start, largely due to the pandemic-driven contraction in 2020.

Jacobsen: Why does the public not track revisions?

Rosner: Messaging is simple; revisions are technical. The benchmark process can significantly change the narrative after the fact, as seen when 2025’s reported gain was revised down from 584,000 to 181,000.

Jacobsen: A few weeks ago, the United States formally withdrew from the World Health Organization. How might that decision affect American workers?

Rosner: The immediate labour-market impact is indirect. Withdrawal from the World Health Organization primarily affects global disease surveillance, coordination, and response capacity. Reduced international cooperation in public health can delay outbreak detection, disrupting supply chains, travel, and economic stability. Those disruptions can eventually affect American workers.

Domestically, the more immediate issue concerns vaccination policy and public health messaging. The United States is currently experiencing its largest measles outbreak in decades. Measles can be fatal in roughly one to three cases per 1,000 infections in developed countries, and it can also cause serious complications such as pneumonia, encephalitis, and long-term immune suppression. Even when it is not fatal, measles can lead to significant health consequences.

Vaccines vary in effectiveness. The measles, mumps, and rubella (MMR) vaccine provides very high protection, typically around 97% after two doses. Other vaccines, such as those for influenza and COVID-19, offer partial protection. They may not prevent every infection, especially with rapidly mutating viruses, but they substantially reduce the risk of severe illness, hospitalization, and death.

Public misunderstanding often stems from the idea that a vaccine must provide 100% protection to be considered effective. In reality, many vaccines are designed to reduce severity and transmission risk rather than eliminate infection.

Leadership and public messaging matter. When officials express skepticism about germ theory or minimize the importance of vaccination, it can reduce public confidence in established medical science. Lower vaccination rates increase the likelihood of outbreaks, which can strain healthcare systems and create broader social and economic consequences. Withdrawal from international public health cooperation, combined with weakened domestic vaccine uptake, poses risks not only to global health but also to national resilience.

Withdrawing from major international public health institutions can weaken global disease coordination. It is difficult to quantify how many lives such policy changes might ultimately affect. However, reductions in surveillance, vaccination coordination, and funding for international response efforts can increase risks worldwide, particularly in lower-income countries that depend on shared resources and data.

Jacobsen: On a related issue, U.S. courts have repeatedly found that Immigration and Customs Enforcement has unlawfully detained individuals. Some analyses suggest thousands of such rulings. What are your thoughts?

Rosner: Courts have, in numerous cases, determined that ICE detentions violated statutory or constitutional standards. When courts find unlawful detention, the key question becomes accountability. If violations are not consistently addressed through internal discipline, civil liability, or, where warranted, criminal prosecution, problematic conduct can persist. Public scrutiny has increased in recent years, which can influence institutional behaviour, but enforcement of legal standards ultimately depends on oversight mechanisms within the Department of Justice and the courts.

There have been high-profile cases in which video evidence later contradicted official accounts of enforcement encounters. In such situations, transparency and independent review are critical to maintaining public trust. Without credible accountability, misconduct allegations can erode confidence in federal law enforcement agencies.

Jacobsen: There have also been debates about federal involvement in election administration and proposals such as the SAVE Act.

Rosner: Under the U.S. Constitution, states administer elections, including federal elections, subject to congressional regulation under Article I, Section 4. Proposals such as the Safeguard American Voter Eligibility (SAVE) Act aim to require documentary proof of U.S. citizenship to register to vote in federal elections. Supporters argue this prevents noncitizen voting; critics argue it could create administrative burdens for eligible voters who lack readily available documentation, such as birth certificates or passports.

Name changes after marriage can create documentation mismatches for some voters, requiring additional paperwork to establish identity continuity. Whether such requirements constitute reasonable safeguards or undue burdens is the subject of ongoing political and legal debate.

As for voter fraud, multiple studies have found that documented cases are rare relative to the total number of ballots cast. Analyses of publicly available databases, including compilations by advocacy organizations, suggest that confirmed fraudulent votes represent a tiny fraction of total votes nationwide. Most experts agree that while election integrity is important, large-scale fraud affecting national outcomes has not been demonstrated in modern U.S. elections.

The reason large-scale voter fraud is rare is straightforward. It is already a felony in most jurisdictions, carrying potential prison time, fines, and a permanent criminal record. The personal risk is high, and the impact of casting a single illegal ballot in an election involving tens or hundreds of thousands of votes is negligible. That cost–benefit imbalance discourages rational actors from attempting it.

Most documented cases tend to involve confusion, clerical errors, or misunderstandings about eligibility rather than coordinated fraud. For example, there have been cases in which individuals with prior felony convictions mistakenly believed their voting rights had been restored. In Texas, Crystal Mason was sentenced to five years in prison in 2018 for casting a provisional ballot while on supervised release; she maintained that she believed she was eligible. Her conviction was later overturned by the Texas Court of Criminal Appeals in 2024. That illustrates how complicated eligibility rules can create legal jeopardy even in the absence of clear criminal intent.

Data compiled by various organizations, including the Heritage Foundation’s Election Integrity Database, lists a few thousand proven fraud cases over several decades, out of billions of ballots cast nationwide. That represents a very small fraction of total votes. Most academic studies similarly conclude that in-person voter impersonation and other forms of individual ballot fraud are rare. Whether new federal legislation is proportionate to that scale is a matter of political debate.

Jacobsen: Turning to foreign policy, Marco Rubio recently visited Europe and framed the United States as historically and culturally connected to its European allies. That tone seems different from the Trump administration’s early posture. Is this a meaningful shift?

Rosner: Public diplomacy often varies from speaker to speaker. Rubio, as Secretary of State, may emphasize shared transatlantic ties and strategic alignment. That rhetoric is not inherently inconsistent with longstanding U.S. foreign policy traditions, which have emphasized NATO cooperation and shared democratic institutions.

However, tone does not necessarily determine policy. Cabinet officials can signal reassurance to allies, but ultimate strategic direction rests with the president. If the administration’s broader policies diverge from conciliatory messaging, allies will evaluate actions more than speeches.

Jacobsen: Some commentators speculate that Rubio appears uncomfortable in his role. Does that interpretation persuade you?

Rosner: It is difficult to infer internal states from public appearances. Photographs or isolated moments rarely provide reliable evidence of personal conviction. What matters is institutional authority and the execution of policy. Regardless of the Secretary of State’s demeanour, the Secretary of State operates within the parameters set by the president. In foreign affairs, structural power outweighs facial expression.

Jacobsen: Let’s move to technology. Elon Musk recently argued that much of the engineering effort behind Tesla’s Optimus humanoid robot is focused on the hands because the human hand is extraordinarily dexterous and has a wide range of motion. What do you think of that claim?

Rosner: That part is reasonable. We talked about this in another session. The human hand is a biomechanical marvel. It has 27 bones, multiple joints per finger, fine motor control, and an opposable thumb that enables precision grips and power grips. Five digits appear to be an evolutionary “sweet spot.” With four fingers and a thumb, humans can manipulate most objects efficiently. More digits would increase mechanical complexity without proportionate functional gain.

Tool use in the animal kingdom shows the distinction. Some birds, such as New Caledonian crows, can shape sticks into hooks using their beaks. That demonstrates intelligence, but hands dramatically expand the range and precision of toolmaking. Dexterous hands plus a large brain create the feedback loop that underlies advanced technology. So if you are building a general-purpose humanoid robot, the hand is one of the hardest engineering problems.

There is also a social component. When robots operate in civilian environments, designers often anthropomorphize them. Even simple delivery robots are given eyes or names because humans respond more comfortably to agents that appear somewhat familiar. A humanoid form with expressive capacity can ease interaction.

Jacobsen: There is also renewed anxiety in the tech sector about the speed of AI progress.

Rosner: That concern is grounded in observable improvement. Generative video systems have recently produced short clips that are increasingly convincing at first glance. A filmmaker named Rory Robinson circulated AI-generated scenes depicting well-known actors in fabricated scenarios to illustrate how quickly the technology is advancing. The point was not to celebrate job displacement but to highlight the potential threat to creative labour.

The larger debate centers on whether AI systems are entering what some call a “phase change.” Earlier systems required close supervision because they generated frequent factual or logical errors. Newer models can perform longer, more coherent tasks with fewer obvious mistakes. That does not make them infallible, but it does reduce the friction involved in delegating substantial work to them.

The economic implication is uncertainty. Historically, automation has replaced some tasks while creating others. The question is whether current AI systems primarily augment human capability or substitute for it. If they approach consistent high-level performance across domains, labour markets could experience meaningful disruption. If their reliability plateaus, they may remain powerful tools rather than wholesale replacements. The trajectory is empirical, not mystical, and it deserves careful measurement rather than panic or denial.

Jacobsen: In that context, how should we define a job? How should we define a career? Is there a third term we should be using?

Rosner: A job is a task bundle for which someone will pay you. A career is a sustained, identity-shaping trajectory built from related job bundles over time, usually with skill accumulation and status progression. There may be a third category emerging: platform-dependent income streams. These are not traditional careers, and often not stable jobs either. They are contingent revenue channels tied to algorithms and audience attention.

Take food delivery. Some roles remain because robotics has not yet scaled sufficiently to replace human labour. That is a job: task-based, replaceable, and often precarious.

Consider digital content platforms such as OnlyFans. OnlyFans reports millions of global creators, with estimates suggesting over one million based in the United States. For context, the U.S. has roughly 800,000 to 1 million sworn law enforcement officers and about 1.1 million physicians. Platform-based adult content creation has therefore become numerically significant, even if income distribution is highly unequal.

However, most creators earn little. A small percentage generate substantial income, often treating the work as full-time, involving marketing, subscriber engagement, and outsourced management services. That resembles entrepreneurship more than traditional employment. It is labour shaped by attention economics.

These platform-mediated roles can offer temporary insulation from AI automation, particularly where human presence, authenticity signalling, or parasocial interaction matter. Yet even these spaces face competition from AI-generated content.

Some sectors attempt formal protection. Entertainment unions such as the Writers Guild of America and SAG-AFTRA have negotiated contract language addressing the use of AI. That signals a shift: workers are not only negotiating wages but also negotiating the boundaries between human and machine contributions.

The larger issue is structural. If AI compresses the demand for cognitive labour, society may need new categories beyond “job” and “career.” We may see more hybrid roles: partial automation supervisors, brand-anchored creators, or trust-based human intermediaries. The labour market may bifurcate into high-skill oversight and low-barrier gig work, with fewer stable middle trajectories.

The definition of work increasingly hinges not on what you do, but on what cannot easily be replicated, automated, or synthetically simulated.

Jacobsen: Entertainment unions have negotiated agreements requiring that a human writer be credited and compensated if a film is produced. But AI systems are improving rapidly. Will there be widespread circumvention?

Rosner: The Writers Guild of America and SAG-AFTRA secured contract language limiting the use of AI as a replacement for credited human labour. The agreements generally state that AI cannot receive writing credit and cannot be used to undermine minimum compensation structures. On paper, that preserves authorship.

In practice, enforcement is more complicated. A technologically literate writer can use AI privately to test ideas, generate alternative scenes, or stress-test plot logic before submitting a script. That resembles research assistance more than outright replacement, but it blurs the line. If AI produces a rough draft and the human substantially revises it, authorship becomes philosophically and contractually ambiguous.

Writers have always used tools: research databases, spellcheckers, and even encyclopedias shaped earlier generations of work. The ethical distinction hinges on the depth of the contribution. Using AI to brainstorm differs from delegating narrative construction wholesale.

Jacobsen: So is using AI now simply due diligence?

Rosner: Increasingly, yes. Many professionals treat AI as an ideation engine or structural critic. It can suggest missed angles, identify inconsistencies, or propose alternate framings. That does not require surrendering authorship. However, if someone relies on AI for a full draft and then lightly edits it, the creative burden shifts. That shift may become common because it is efficient.

The economic parallel resembles discount marketplaces. Just as surplus bakery goods can be redistributed at lower cost through apps, cognitive surplus can be redistributed through AI tools. High-end handcrafted work still exists, but cheaper approximations proliferate. Consumers may not always distinguish between them.

Jacobsen: Does this mean professional writing is sliding toward assisted automation?

Rosner: It likely becomes hybrid. Purely human writing will persist where originality and voice are valued. But baseline commercial writing—marketing copy, formulaic scripts, procedural content—may increasingly begin with machine drafts refined by humans.

The critical variable is transparency. If AI is treated as a tool, like a calculator for language, the profession adapts. If it becomes an invisible ghostwriter, attribution norms destabilize. The trajectory depends less on capability alone and more on cultural standards about what counts as authorship.

You’ve described this as a mixed landscape. Do you think we are moving through a genuine phase shift in labour?

Jacobsen: It looks transitional, not apocalyptic. In prior industrial revolutions, machines automated physical labour, and new industries emerged around that automation. This wave differs because it targets cognition. We are not only mechanizing muscle; we are partially automating pattern recognition, drafting, diagnosis support, and analytical synthesis.

That does not mean total replacement. It means uneven impact. Some roles will be lightly assisted. Others will be substantially reduced. A few may disappear. The distribution may resemble a gradient rather than a binary outcome. Tasks cluster along a spectrum from augmentation to substitution.

Rosner: Take paralegals. In a large firm employing several, AI-assisted document review and drafting tools could reduce staffing needs. A firm that once required three paralegals may function with one or two. In a small practice with limited resources, AI could eliminate the need to hire entirely for certain routine tasks. That does not eliminate legal work, but it compresses support roles.

Tax preparation offers a parallel. For straightforward returns, automated platforms already handle filing without a human accountant reviewing each case. Complex returns still require certified professionals. AI shifts the threshold of complexity at which human expertise becomes necessary.

Medicine illustrates the hybrid phase. Clinical documentation consumes a significant share of physician time, sometimes estimated at around one-quarter to one-third of work hours. AI transcription and drafting systems can reduce that burden. However, hallucinations and factual errors remain risks and require oversight. In clinical settings, even small inaccuracies carry liability consequences. So the human remains the verification layer.

The direction of travel suggests that supervision requirements will decrease as models improve. Yet even if AI reduces paperwork from 30 percent of a physician’s day to 20 percent, that is augmentation, not displacement. The more sensitive the domain—law, medicine, finance—the longer humans remain embedded as accountability nodes.

The deeper structural question is whether fully automated cognitive loops can generate new industries or merely consolidate existing ones. If AI both performs and supervises certain industrial tasks, new job creation may not scale proportionally to the displacement it causes. That is what distinguishes this phase from earlier revolutions.

Still, history counsels caution against deterministic predictions. Intelligence automation is powerful, but institutions, liability systems, trust structures, and human preferences slow total substitution. 

Jacobsen: The landscape ahead is neither a clean bell curve nor a collapse. It is a layered terrain of partial automation, human oversight, and evolving definitions of expertise. Do you suspect many physicians are already using AI for documentation?

Rosner: I assume a significant number are experimenting with AI-assisted transcription and note drafting. Electronic medical record systems increasingly integrate voice-to-text and structured summarization tools. The efficiency gain is real, but so is the need for verification. In medicine, a small hallucinated detail can have legal and clinical consequences. So adoption tends to be cautious and layered with oversight.

Jacobsen: Thank you very much for the opportunity and your time, Rick.

Rick Rosner is an accomplished television writer with credits on shows like Jimmy Kimmel Live!, Crank Yankers, and The Man Show. Over his career, he has earned multiple Writers Guild Award nominations—winning one—and an Emmy nomination. Rosner holds a broad academic background, graduating with the equivalent of eight majors. Based in Los Angeles, he continues to write and develop ideas while spending time with his wife, daughter, and two dogs.

Scott Douglas Jacobsen is a collaborator with Rick Rosner. He is the publisher of In-Sight Publishing (ISBN: 978-1-0692343) and Editor-in-Chief of In-Sight: Interviews (ISSN: 2369-6885). He writes for The Good Men Project, International Policy Digest (ISSN: 2332–9416), The Humanist (Print: ISSN 0018-7399; Online: ISSN 2163-3576), Basic Income Earth Network (UK Registered Charity 1177066), A Further Inquiry, Washington Outsider, Rabble, and other media. He is a member in good standing of numerous media organizations.

Photo by Imagine Buddy on Unsplash

Last updated May 3, 2025. These terms govern all In Sight Publishing content—past, present, and future—and supersede any prior notices. In Sight Publishing by Scott Douglas Jacobsen is licensed under a Creative Commons BY‑NC‑ND 4.0; © In Sight Publishing by Scott Douglas Jacobsen 2012–Present. All trademarks, performances, databases & branding are owned by their rights holders; no use without permission. Unauthorized copying, modification, framing or public communication is prohibited. External links are not endorsed. Cookies & tracking require consent, and data processing complies with PIPEDA & GDPR; no data from children < 13 (COPPA). Content meets WCAG 2.1 AA under the Accessible Canada Act & is preserved in open archival formats with backups. Excerpts & links require full credit & hyperlink; limited quoting under fair-dealing & fair-use. All content is informational; no liability for errors or omissions: Feedback welcome, and verified errors corrected promptly. For permissions or DMCA notices, email: scott.jacobsen2025@gmail.com. Site use is governed by BC laws; content is “as‑is,” liability limited, users indemnify us; moral, performers’ & database sui generis rights reserved.