Diablo Valley Employers Should Pay Students to Check AI’s Work

Artificial intelligence is changing what employers expect from workers. As AI tools become capable of producing reports, analyzing documents, and recommending decisions, a new workplace skill is becoming increasingly important: knowing when the technology is wrong.

For students preparing to enter the workforce, learning to use AI is only part of the challenge. They must also learn to question its conclusions, identify errors, and explain why an answer should or should not be trusted.

Diablo Valley College and local employers have an opportunity to make those skills part of practical, paid work experience.

On October 8, Diablo Valley College hosted its Business, Technology, and Engineering Fair, connecting students with employers and organizations offering career opportunities. The event reflects the college’s ongoing efforts to connect education with the changing needs of the workplace.

DVC is also preparing students for careers involving artificial intelligence. The college has offered an Introduction to AI Programming course and has described plans for a certificate in AI Programming, tentatively targeted for 2027.

These efforts are valuable. But the next opportunity may be less about teaching students to create AI systems and more about teaching them to evaluate the work those systems produce.

AI creates a new responsibility for workers

The growing use of AI does not eliminate the need for human judgment. In many situations, it makes that judgment more important.

An AI system can generate a convincing financial summary while overlooking an important qualification. It can produce a customer-service response that sounds professional but misinterprets company policy. It can compare suppliers while relying on incomplete or outdated information.

The problem is not simply that AI makes mistakes. People make mistakes, too.

The challenge is that AI can present incorrect conclusions with the same confidence and polish as correct ones. Employees who become accustomed to accepting those outputs may gradually lose the habit of checking the underlying evidence.

Recent research illustrates why these skills matter.

In an October 16 report examining how employment is changing at Anthropic, researchers described how increased AI adoption has altered the company’s work organization, productivity, and hiring priorities.

The report found that employees increasingly spend time overseeing AI-generated work and managing automated processes. It also described a growing emphasis on autonomy and judgment in hiring decisions.

Anthropic is an unusually AI-intensive workplace, and its experience should not be assumed to represent every business. Nevertheless, the research highlights an important question for employers everywhere: Who will evaluate the growing volume of work produced by AI systems?

That responsibility should not fall exclusively on experienced managers.

Students entering the workforce can begin developing these skills before graduation.

A paid verification pilot for Diablo Valley students

Local employers could collaborate with interested students and educational institutions to create a small paid AI-verification program.

The idea would be straightforward.

Participating businesses would provide supervised assignments involving AI-generated work. Students would examine the original information, identify discrepancies, document supporting evidence, and explain when a decision requires additional human review.

Consider a local business using AI to compare potential suppliers.

The system produces a polished recommendation identifying the least expensive option. But the comparison overlooks a delivery restriction that could make the recommended supplier unsuitable.

A student assigned to review the analysis would need to examine the original supplier information, identify the missing condition, and explain how it changes the recommendation.

Another employer might ask students to review AI-generated customer-service summaries for unsupported promises or inaccurate interpretations of company policies.

A nonprofit organization could use appropriately protected records to teach students how to identify unsupported claims in AI-generated reports.

The assignments would not require students to make consequential business decisions independently. Their role would be to verify information, explain their findings, and escalate questions to qualified supervisors.

Businesses would retain responsibility for reviewing and approving the final work.

Start small and measure the results

An initial pilot could involve three to five participating employers and six to ten students over six to eight weeks.

Each student might complete four to six paid hours of supervised verification work per week.

These figures are illustrative rather than established program requirements. No participating employers or institutional commitments have been announced.

The pilot could focus on three practical abilities.

First, students would learn to identify factual errors and unsupported conclusions in AI-generated work.

Second, they would learn to document their reasoning using original source material rather than relying on the AI system’s explanation.

Third, they would practice recognizing uncertainty and knowing when to ask for additional guidance.

Employers could evaluate the results by measuring how accurately students identify errors, how clearly they explain their findings, and whether they recognize situations requiring escalation.

Assignments should use fictional, public, or appropriately protected information, with clear supervision and confidentiality requirements.

Students should receive compensation for productive work. A program intended to prepare young people for employment should recognize the value of their time and contributions.

A different kind of workforce preparation

Teaching students to verify AI-generated work would complement, rather than replace, technical AI education.

Some students will pursue careers building AI systems. Others will enter accounting, healthcare administration, marketing, logistics, public service, and other fields where they may use AI without ever developing the underlying technology.

Those workers will still need to understand the limitations of the systems they rely on.

A marketing employee should be able to recognize when AI invents a statistic. A purchasing assistant should notice when a supplier comparison excludes important conditions. An administrative worker should question a summary that contradicts the original document.

These are not exclusively technical skills. They involve evidence, reasoning, communication, and professional responsibility.

They are also skills that can be taught and evaluated through practical assignments.

Diablo Valley College already provides career education designed to connect classroom learning with employment opportunities. A paid AI-verification initiative could build on that approach while giving participating businesses a practical way to explore responsible AI adoption.

Such a program would require interested employers, appropriate institutional coordination, supervision, and a clear compensation structure. It should begin as a limited experiment rather than a promise of a new workforce model.

But it is an experiment worth considering.

The future workplace will need people who can work productively with artificial intelligence. It will also need people who recognize when a convincing answer is incomplete, misleading, or simply wrong.

Local employers have an opportunity to help students develop both abilities.

And they should be willing to pay students for the work of learning to tell the difference.

---

About the author

Gleb Tsipursky, PhD, is a behavioral scientist, CEO of Disaster Avoidance Experts, and author of The Psychology of AI Adoption at Work: From Resistance to Results (Georgetown University Press, 2026). His writing on organizational decision-making and AI adoption has appeared in Harvard Business Review and other publications.

Editor’s note: This article presents the author’s proposal for a potential paid student-employer initiative. The proposed pilot is not an existing Diablo Valley College program, and no partnership or employer commitment has been announced. The views expressed are those of the author.