Construir tu primer bot o asistente personalizado

Item: Construir tu primer bot o asistente personalizado
Rating: 4
Author: PostSingularity Institute

Build your first custom bot or assistant

by Habilidades de Inteligencia Artificial Aplicada Uk · View the channel on PSI

Applied Sciences & Engineering AI — Models & Research UYQArtificial intelligence (AI)UYQLNatural language and machine translation

🎙 Habilidades de Inteligencia Artificial Aplicada Uk 👥 3K 📅 October 17, 2026 ⏱ 106 min 👁 19 📄 tutorial 🧭 2026-10-18

Available in: English (current) Français

Keywords

AI assistantGeminiSystem promptCustom botPrompt design

Summary

This video is a comprehensive lecture on creating custom AI assistants (bots) using Google’s Gemini. The instructor begins by recapping the previous session, where attendees built personalized assistants for specific tasks. The focus shifts to designing a bot that other people can use without further explanation, emphasizing clarity and precision in instructions. The core of the session is about crafting a high-quality system prompt, which is compared to an operation manual for the AI. Four essential components are detailed: role, objective and tone, response structure, and boundaries. The instructor provides examples, such as an email assistant, and engages the audience through interactive questions and discussions. Participants share their own prompt ideas, including sales support, educational planning, and dental appointment scheduling. The session includes an activity where participants write their own system prompts, followed by peer feedback. The instructor highlights the importance of anticipating out-of-scope requests and vague queries. The video concludes with guidance on implementing the prompt in Gemini, emphasizing iterative improvement. Overall, it serves as a practical tutorial for individuals seeking to build reusable AI assistants with clear, well-defined behavior.

181 words

Critical Evaluation

Value of the Information & Strength of the Argument

The value of this video lies in its practical, hands-on approach to teaching AI assistant design. The instructor systematically breaks down the creation of a system prompt into four actionable components, providing concrete examples and encouraging audience participation. The argumentation is based on experiential learning and common-sense principles, rather than formal research or citations. While this approach effectively conveys the core concepts, it lacks empirical evidence or comparative analysis to support claims about prompt effectiveness. The interactive format strengthens engagement but may dilute the depth of theoretical explanation. Nonetheless, the tutorial-like structure makes it highly accessible and immediately applicable for viewers.

Scientific Rigor, Source Quality, Title Accuracy

Scientifically, the video does not present original research or cite academic sources. It relies on general AI knowledge and the presenter’s personal experience. The absence of references limits its value for academic purists, but for practical application, the content is clear and logically structured. The title accurately matches the content, focusing on building a custom bot with Gemini. There are no external sources in the description, and no citations are mentioned in the transcript. The presentation is coherent and pedagogically sound, though it does not engage with potential limitations or alternative approaches in detail.

210 words

Title / Content Match

Title accurately reflects content: building a personalized bot/assistant using Gemini and system prompts.

Quality & Reliability

7/10

Clear pedagogical structure, accurate information about system prompts and Gemini, but lacks rigorous sourcing and scientific depth.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Contribution & Novelties

The originality of this video lies in its structured pedagogical approach to teaching system prompt design for custom AI assistants. It provides a clear framework that helps novices understand how to specify role, tone, structure, and boundaries to make an assistant reusable by others. The emphasis on anticipating out-of-scope requests and vague inputs is particularly practical.

Pour aller plus loin :

Prompt engineering — Background on techniques for crafting AI instructions.
System prompt — Detailed explanation of system prompts in AI models.
Chatbot — General overview of chatbot technology and applications.
Gemini (language model) — Information about Google’s AI model used in the tutorial.

104 words

Radar Profile

The radar profile shows balanced scores with relatively high information quantity (8) and quality (7), but lower technical depth (6) and reliability (7). This indicates a verbose, accessible tutorial with solid practical value, though it lacks rigorous scientific grounding.

Reliability 7/10