We Were Wrong About AI All Along…

Overview

The presenter reflects on how AI is already automating routine, screen-based work. Drawing on experience in consulting, purchasing, logistics, planning, administration, and finance, he says he has used AI in his own business to automate tasks he previously completed manually.

He argues that office work is especially exposed because much of it involves digital information, coordination, and checking that systems are functioning. The key distinction is not between “intelligent” and less-skilled work, but between work that can be done entirely on a computer and work that requires physical presence in unpredictable environments. AI may help diagnose, plan, or monitor a job, while hands-on repairs remain difficult to automate with current robotics.

Skilled Trades and Business Opportunities

The presenter identifies five skilled trades that may be harder to automate fully:

• Electricians: Demand continues across homes, buildings, data centers, and electric-vehicle charging. The work can involve troubleshooting unfamiliar or outdated wiring.
• Plumbers: AI can explain systems or assist with diagnosis, but repairs in confined spaces and undocumented installations require on-site work.
• Elevator maintenance technicians: The job combines mechanical, electrical, safety, and software systems, often in cramped spaces.
• Power-line installers and lineworkers: Restoration work may involve heights, bad weather, fallen trees, and damaged equipment.
• HVAC technicians: AI can monitor temperatures and predict failures, but broken equipment still requires physical repair. The presenter notes the irony that AI data centers also need cooling.

These careers are not risk-free: they can be physically demanding or dangerous and require specialized training. The presenter also emphasizes that trades can offer a path to self-employment or to building a company over time.

Business and Career Frameworks

• Digital-work exposure audit: Review a typical workday and focus on actual tasks rather than job title. Consider how much time goes to emails, spreadsheets, meetings, tracking, and verification—and how much requires physical presence. Work that happens entirely on-screen may be more accessible to software.
• “Multiple pillars” resilience framework: Rather than asking which single job is secure, consider how dependent you are on one employer, one salary, and one plan for your life. Developing projects or income sources alongside employment can reduce dependence on a single source.
• Human-plus-AI workflow: AI may help tradespeople work faster by supporting diagnosis, planning, monitoring, or failure prediction. However, higher productivity does not guarantee higher pay or continued staffing; businesses may also use productivity gains to reduce headcount.
• Operational insight: Some routine checks and follow-ups exist because systems are not working well. The presenter suggests AI is particularly capable of improving or automating these processes.

Examples and Actionable Takeaways

• The presenter’s own use of AI in his business demonstrates that automation is already practical for some manual tasks.
• When his household lost power, it took about two days for an electrician to arrive. The experience illustrated how dependent people can be on specialized, locally available services.
• He repaired a broken washing machine himself after researching how it worked, illustrating the gap between accessing information and carrying out physical work.
• A practical first step is to map the previous day’s work, identify which tasks are fully digital, and assess how much the job depends on a single employer or income source.

Metrics and Timelines

• The presenter says he has observed AI agents entering workplace systems over the past two years.
• The power-outage example involved a two-day wait for an electrician.
• No revenue, margin, customer-acquisition, salary, or other standard business KPIs or targets are provided.

Presenter/source: Tom Scryleus.