Will Our 2026 Graduates Lead AI Agents or Lose Jobs to Them?

Students won’t be just competing against AI agents for jobs, they’ll also be competing against new graduates who know how to manage AI agents in the ways described. Who will get the job? :)

Many write about the arrival of AGI. I obviously do a lot, and I think it’s an important concept to think about. But it’s not just because we will then have human-level intelligence across all domains, but because it provides a window to think about what will arrive along the way.

As was forcasted in the spring of 2023, what we are seeing now is the arrival of AI agents.

Basically, AI Agents are AIs that can be given goals, reason as to how to accompish the goals, choose tools to help them complete the goals, and then execute, evaluate, and iterate. And like us, they will be able to do at least many significant parts of at least entry level jobs and really any job that is, very broadly speaking, repetitive.

Some additional examples —

Yesterday, I reported on coding agents. Joshua Clear added some more detail today: COXDEX-1 (OpenAI) assistants simultaneously manage 3-5 development tasks—including bug fixes, test creation, and feature implementation—each running in isolated cloud environments. This represents the next evolution in DevOps—Codex Agents deliver complete working commits rather than just code suggestions. The system processes entire repositories, launches containers to test different approaches, and generates ready-to-implement pull requests. Effectively functioning as team members rather than just tools, these agents are increasingly handling routine development tasks like unit testing, localization, and architectural refactoring.

And we are already seeing gains beyond coding. Flowith can reportedly complete 1,000 sequential tasks without human intervention.

These agents clearly will play a big role in the (immediate) future.

Yesterday’s Googe I/O Conference had a near endless discussion of AI agents operating across the Gemini platform.

They are still somehat limited (they can only complete a limited number of steps, probably up to 8 hours in some unique circumstances), but substantial gains are expected by the end of the year.

Next year’s June 2026 graduates will enter the job market and be among very capable AI agents that will be trained to do much of what they went to school for. Freshman entering school in 2025 will compete against agents in 2029 that are likely millions of times more capable than them.

The question then becomes: Will the students have to compete against the AI agents or manage them?

Students today face a fundamental shift in how they'll need to engage with AI. Rather than competing directly against AI systems, which are increasingly outperforming humans in specific domains, students should be prepared to manage and collaborate with these powerful tools.

Why Direct Competition with AI Is Challenging

1. Speed and scale advantages: AI can process vast amounts of information and perform calculations far faster than humans.

2. Memory capabilities: AI systems can store and instantly recall enormous datasets without the limitations of human memory.

3. Consistency and endurance: AI doesn't experience fatigue, emotional fluctuations, or attention lapses that affect human performance.

4. Specialized optimization: Many AI systems are designed to excel at specific tasks, often surpassing human capabilities in those narrow domains.

5. Rapid advancement: The capabilities of AI systems continue to improve at a pace that humans cannot match biologically.

Training Students to Become Effective AI Managers

Instead of competing against AI, students should learn to:

1. Develop AI Literacy and Understanding

• Learn the fundamentals of how AI systems work, their capabilities, and limitations

• Understand different types of AI and their appropriate applications

• Recognize AI biases and how to mitigate them

2. Focus on Human Skills

• Develop creativity and original thinking

• Build emotional intelligence and interpersonal skills

• Exercise ethical judgment and values-based decision making

• Practice contextual understanding and cultural awareness

3. Learn Effective Human-AI Collaboration

• Understand when to rely on AI and when human judgment is superior

• Design workflows that combine human and AI strengths

• Communicate effectively about AI capabilities with non-technical stakeholders

4. Build Technical Integration Abilities

• Learn to connect different AI tools into effective workflows

• Develop basic programming skills to customize and extend AI capabilities

• Understand data preparation and management for AI systems

Training Students to Become Effective AI Managers: Marketing Major Example

For a marketing major, effective AI management will be crucial as AI increasingly transforms how marketing campaigns are designed, implemented, and measured. Here's a detailed example of how a marketing student might develop AI management skills across various marketing domains:

Semester-Long Marketing Campaign Project

Phase 1: Campaign Planning & Research

AI Management Task: Directing market research and consumer insight gathering

The student would:

• Create targeted prompts to have AI analyze demographic data and identify market segments

• Direct AI to search for relevant consumer behavior trends across different platforms

• Evaluate AI-generated consumer personas, refining those that lack nuance or contain stereotypes

• Use AI to rapidly generate multiple campaign concepts based on research findings

• Critically assess which AI-generated ideas align with brand values and target audience preferences

Example Prompt Sequence:

1. "Analyze this dataset of consumer behavior for adults 25-34 in urban areas and identify the top 5 purchase motivations for sustainable products."

2. "Now cross-reference these motivations with social media conversation trends from the past 6 months."

3. "Based on these insights, generate 3 distinct consumer personas that represent different segments within this demographic."

The student would then critically evaluate these personas, identify gaps or unrealistic elements, and direct the AI to refine them.

Phase 2: Content Development

AI Management Task: Orchestrating content creation across channels

The student would:

• Define brand voice guidelines for the AI to follow when generating content

• Direct AI to develop variations of ad copy for A/B testing

• Instruct AI to adapt messaging across platforms while maintaining brand consistency

• Review AI-generated content for authenticity and emotional resonance

• Combine human creative direction with AI implementation for visual assets

Example Workflow:

1. The student creates a detailed creative brief including target audience, key messages, and emotional tone

2. They instruct AI to generate 10 headline options following these guidelines

3. After selecting promising options, they direct AI to expand these into full ad copy variants

4. They critically assess the AI-generated content, identifying which pieces feel authentic vs. generic

5. They provide specific feedback to refine the most promising content: "The emotional appeal in version B feels more authentic, but needs to emphasize our sustainability message more clearly"

Phase 3: Campaign Optimization

AI Management Task: Using AI for data analysis and campaign refinement

The student would:

• Set up parameters for AI to monitor campaign performance metrics

• Direct AI to identify patterns and correlations in customer engagement data

• Evaluate AI-recommended optimization strategies against business objectives

• Make informed decisions about resource allocation based on AI insights

• Translate technical analytics into actionable marketing adjustments

Example Scenario: The campaign has been running for two weeks. The student directs the AI to:

1. Analyze performance data across all channels

2. Identify which demographic segments are responding best to which messaging

3. Suggest three specific adjustments to improve performance with underperforming segments

4. Project the potential impact of these adjustments based on current response patterns

The student then evaluates these suggestions, considering factors the AI might not fully grasp (like brand positioning concerns or competitive landscape shifts), and implements a revised strategy.

Phase 4: Reporting & Strategic Planning

AI Management Task: Creating comprehensive analysis and future recommendations

The student would:

• Direct AI to compile campaign results into a cohesive narrative

• Critically evaluate AI-generated insights and their business implications

• Use AI to model different scenarios for future campaigns

• Develop a presentation that combines AI-generated data visualization with human strategic thinking

• Create a knowledge management system for insights that can be applied to future campaigns
  
  Now, students won’t be just competing against AI agents for jobs, they’ll also be competing against new graduates who know how to manage AI agents in the ways described.
  
  How do you think will get the job? Do you think it will go to the student who has been taught that using AI is cheating?