Make an AI trends presentation
Manus AI has been making waves as a general-purpose AI agent that can do just about anything — browse the web, write code, analyze data. Task: The author tested Manus 1.6 Lite to generate a presentation on "AI Trends in 2026" using a specific prompt, comparing it against PageOn.ai. The workflow involved initial generation followed by a manual edit command to add charts. Manus performed deep web research, citing sources like Deloitte and Stanford HAI, but generated slides without pausing for user confirmation on the outline.; tools: Manus. Human role: The original post does not say how a person stayed involved.; limitations: Manus lacks an outline confirmation step, forcing users to accept AI assumptions about content emphasis. Visual output is limited to text and basic blocks, requiring significant manual redesign for client-facing presentations. The tool is better suited for internal research briefs where visual polish is secondary to source credibility and content depth..
Tools
Published on Agent Work Atlas:
Overview
Manus AI has been making waves as a general-purpose AI agent that can do just about anything — browse the web, write code, analyze data.
Workflow
The author tested Manus 1.6 Lite to generate a presentation on "AI Trends in 2026" using a specific prompt, comparing it against PageOn.ai. The workflow involved initial generation followed by a manual edit command to add charts. Manus performed deep web research, citing sources like Deloitte and Stanford HAI, but generated slides without pausing for user confirmation on the outline.
Reported result
Manus produced an 11-slide deck with strong content depth but poor visual variety, featuring monotone layouts and zero images. The research process was transparent and traceable, but the generation took over 10 minutes. A subsequent edit to add charts took another 7-8 minutes and resulted in only three basic visualizations, highlighting inefficiency for iterative design.
Human role
The original post does not say how a person stayed involved.
Limitations
Manus lacks an outline confirmation step, forcing users to accept AI assumptions about content emphasis. Visual output is limited to text and basic blocks, requiring significant manual redesign for client-facing presentations. The tool is better suited for internal research briefs where visual polish is secondary to source credibility and content depth.
Try it
Act as a presentation designer using Manus 1.6 Lite. I need a draft presentation on [replaceable topic, e.g., 'AI Trends in 2026'] for [replaceable audience, e.g., 'internal tech team']. Step 1: Input the topic into Manus and monitor the real-time research process, noting key sources like [replaceable source examples, e.g., 'Deloitte Tech Trends']. Step 2: Wait for the initial outline generation. Since Manus does not pause for confirmation, immediately review the outline for relevance. Step 3: If the outline is acceptable, allow slide generation. Step 4: After generation, issue the command 'add charts to wherever it needs' to enhance data visualization. Step 5: Export the slides and manually redesign the visual layout to include images and varied templates, as Manus output is text-heavy. Measure success by verifying that all cited data points are traceable to the research logs and that the final deck meets visual standards after manual editing. Do not publish or present the raw Manus output without human review of both content accuracy and design quality.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.