(Never underestimate an old man who appear to pretend like he's an idiot)

Please revise the YouTube advertisement that suggests people who do not learn AI will be left behind in employment. At the same time, it is important not to underestimate Microsoft Excel, as it is capable of handling complex calculations where even AI can sometimes make errors. For someone who has experience using multiple AI applications and has also lived through two different technological eras, I can say from experience that I understand both perspectives.

Later someone challenge me to provide an example. CHALLENGE ACCEPTED!! :

One common example is multi-layered engineering or financial spreadsheets with interdependent formulas, lookup tables, conditional logic, and iterative calculations.

For instance, in a construction project cash flow model:

• Multiple worksheets are linked together,
• Progress claims are calculated based on work completed,
• Retention sums, liquidated damages, variations, taxes, escalation factors, and currency conversions are applied,
• Formulas may contain nested IF, INDEX-MATCH, XLOOKUP, SUMPRODUCT, and array functions,
• Circular references may be intentionally used with iterative calculations.

An AI can often:

• Misinterpret the logic flow between worksheets
• Miss hidden cells, named ranges, or indirect references
• Suggest formulas that appear correct but produce inaccurate results.
• Fail to detect that a small formula change affects dozens of downstream calculations.

A simpler but famous example is: =SUMPRODUCT((A2:A100="Approved")*(B2:B100>50000)*(C2:C100)) Many AI systems can explain what the formula does, but when asked to recreate a more complex version involving multiple criteria, date ranges, exclusions, and dynamic arrays, they sometimes generate formulas with subtle errors that return incorrect totals. In engineering, AI also struggles with spreadsheets that combine: - Structural load calculations, - Reinforcement schedules, - Hydraulic flow computations, - Pump sizing calculations, - Earthwork volume balancing, where the spreadsheet creator has embedded years of domain knowledge into the formulas. A fair statement would be: AI is an excellent assistant for spreadsheet development and troubleshooting, but it is not infallible. Complex Excel models built by experienced engineers, accountants, analysts, and project controllers often contain layers of business and technical logic that AI may misunderstand or oversimplify, leading to calculation errors if its outputs are not independently verified. This is why many professionals view AI as a productivity tool rather than a replacement for expertise. Excel remains one of the most powerful analytical tools ever created, particularly in the hands of someone who understands both the formulas and the underlying subject matter.

 

(Never underestimate an old man who appear to pretend like he's an idiot)