Video summary

HAPUS CHATGPT & GEMINI?! | Belajar AI dari 0 di 2026

Main summary

Key takeaways

Technology

Summary of Technological Concepts & Product Features (AI / “Cloud” vs GPT / Gemini)

Overall thesis

The speaker argues that AI platforms can realistically replace parts of human work. They also claim their own workflow has shifted from ChatGPT to an AI product referred to as “Cloud” (often compared against GPT and sometimes “Claude/Claud.Aaii”).

Why “Cloud” over ChatGPT (daily use)

The speaker describes “Cloud” as:

  • More friendly
  • Better for brainstorming
  • Critically, constantly gaining features, implying a fast iteration cycle

Major value proposition

“Cloud” is positioned as a “digital brain”—an operational system for work rather than only a chat interface.

Key “Cloud” Features Mentioned (with practical workflow implications)

1) Import / prompt portability + reduced repetition

In Settings → Capabilities, users can import prompts/instructions from another AI “folder” into Cloud to reuse memory and work context.

Goal: avoid rewriting the same preferences and instructions across tools.

2) Persistent personal instructions

Cloud supports persistent “expert mode” preferences, such as:

  • Double-checking outputs
  • Being skeptical and doing research
  • Optimizing for accuracy
  • Reminding the assistant that it isn’t always right

These function like system-like instructions that carry forward into future outputs.

3) Built-in workspace organization

The workflow includes structured workspace elements such as:

  • New chat
  • Search chat
  • Customize
  • Customize apps/tools integration (e.g., Canva, Google Drive, and others)

It also highlights a Skills ecosystem:

  • Skills are add-ons that expand capability
    • Examples mentioned: Canva design, brand guideline workflows
  • Skills shared by others can be explored via links in the description.

4) Model lineup and recommendation by user level

Models referenced include Haiku, Sonnet, and Opus (plus “tons of other models”).

Recommendation described:

  • Beginners: start with Haiku
  • More capable / extreme use cases: use Opus after upgrading

The speaker claims differences are minimal on free tiers, but become more significant on paid plans, and they use a desktop app + Pro.

5) “Cowork” (team collaboration / shared AI workspace)

“Cowork” is described as a way to chat together with a team inside Cloud. Features include:

  • Schedules (e.g., daily briefing at a set time)
  • Projects to categorize work areas (e.g., content/account tasks by brand)
  • Folders attached to a project so relevant files are used
  • A structured “command center” style page for coordination

6) Automation via schedules + recurring reporting

Example: a daily briefing at 9:00 a.m. about news (e.g., “Teag AIox News Indonesia”).

Core idea: automation generates a timed update without needing manual prompts.

Review / Demo-Style Use Cases the Speaker Demonstrates

A) Comprehensive social media account analysis

The speaker runs analysis for an Instagram/TikTok presence, covering:

  • Plan/steps: review profile, scan captions/hashtags, check story usage
  • Engagement/follower ratios
  • Content mix
  • Audience distribution
  • Recommendations, such as:
    • Diversify content proportions
    • Launch a signature series
    • Build an email list
    • Explore long-form content

Claimed benefit: reduces the “lazy” manual effort of monthly reporting for agencies.

B) “One-shot” creation of landing pages / page designs

The speaker uses a single prompt to generate:

  • A landing page UI
  • Preview and copy/embed capability

They mention inserting the output into a landing page workflow using a URL-like mention such as “lnk.tid,” then editing as needed.

C) Content repurposing: turning transcripts into slide decks

Feeding a YouTube transcript into Cloud produces:

  • 7 Instagram carousel slides

They also iterate on styling (e.g., “Apple style,” cleaner design tone).

D) “Remotion” / video motion graphic attempt

The speaker attempts a Remotion-related skill. When errors occur:

  • Cloud provides a feedback loop
  • The user shares a screenshot/error
  • The assistant guides what’s okay/not okay

E) “Command center” for middle management and team visibility

The speaker demonstrates generating a structured team operating page with:

  • Daily objectives
  • Role descriptions (e.g., CEO, community manager, production supervisor, content admins/editors)
  • Links to deliverables and status/counts

Claimed effect: improved coordination/visibility for middle management and clearer ownership for staff.

F) “Vibe Coding” / app building from prompts (front-end preview)

The speaker showcases prompt-driven UI/app generation using an example “Habit 30” app with:

  • Gamification (XP, levels, badges)
  • Calendar and stats

They emphasize speed: idea → app prototype appears immediately in preview.

They also claim that, behind the scenes, it’s not “just HTML,” and that it may reduce MVP timelines from months to about ~1 month.

Comparisons and Cautions

Compared vs ChatGPT

“Cloud” is claimed to have:

  • A better UI
  • More integrations
  • More workflow automation (not only chat)

Gemini / other services

The speaker mentions:

  • Gemini / Google LM
  • Perplexity
  • Claude

They suggest performance varies by task (e.g., research vs generation).

Meta caution about the AI ecosystem

The speaker notes founder/media dynamics and encourages independent research due to possible propaganda/uncertainty in the AI space.

Main Speakers / Sources (as mentioned)

  • Primary speaker/narrator: “Gusto Halim” (the speaker refers to himself and mentions his channel/audience)

  • Referenced person/source: Seth Godin (mentioned as recommending “Cloud”-type thinking/friend metaphor via a podcast with “Feris”)

Original video