Page 50 - CyberSecurityEssentialsEnglish
P. 50
judgment that is difficult to measure? State that part explicitly in the workflow rather
than assuming everyone knows it.
● Remember: Section summary
• Key idea: The best starting point is a small experiment on a real task, with clear
measures and representative cases.
• Practical decision: Base adoption on reliable benefit and the impact of error, not on the
most impressive example.
• Mistake to avoid: Moving from one successful trial to broad adoption or full automation
without testing exceptions.
Chapter Conclusion: Understand Before You Use
Artificial intelligence is not a single tool. It is a field that includes rule-based systems, systems
that learn from data, and systems that combine both approaches. AI applications can
classify, forecast, rank, generate content, and propose plans. Learning patterns, however,
does not automatically become human understanding or a guarantee of truth.
We have seen that language models generate text by predicting what comes next in context.
They can therefore write fluently and still produce a convincing error. We have also seen
that a system may succeed at one task and fail at another that appears similar. A model’s
reputation is no substitute for testing it on realistic cases.
Practical value appears when technology accelerates preparation, broadens the range of
alternatives, and executes repeated steps under clear conditions. People remain responsible
for the objective, standard, review, and decision. With a clear understanding of AI’s
capabilities and limits, the next question is: how do we communicate with it in a way that
turns an idea into a clear task and a professional result? That is the subject of Chapter Two.
Chapter Glossary
TABLE 1-11: GLOSSARY OF ESSENTIAL TERMS
Term Common abbreviation Plain-language explanation
A scientific and technical field
concerned with systems that perform
Artificial intelligence AI
tasks such as recognition, analysis,
learning, and decision support.
A program or arrangement that
AI system AI system receives data or instructions and
produces an output for a specific task.
A sequence of steps or rules for solving
Algorithm -
a problem or performing a task.
49
