Completing the journey from a storyboard script to four 3D storyboard images, voiceover, subtitles, and a vertical MP4 in 14 minutes was a valuable AI automation test. But the final video could not play smoothly, showing that “able to execute” remains a long way from “able to deliver reliably.”
We noticed that this task did more than ask a model to write copy. It asked the model to plan and connect image generation, voiceover, subtitling, video assembly, and file archiving. Tencent WorkBuddy, a desktop AI workflow tool from Tencent, also automatically installed the missing edge-tts voice tool, located the subtitle-encoding issue, and repaired it. This demonstrates AI’s shift from answering questions to operating computer tools.
What this is
The test covered 18 steps. The author entered three sets of natural-language instructions, and the system generated four 3D cartoon storyboard images, produced a 15.8-second Chinese voiceover, added subtitles and a push-and-pull camera movement, then exported a 13.7 MB vertical MP4 and archived the assets in a local folder.
Unlike an ordinary chatbot, WorkBuddy can read and write local files, execute code, install environments, and invoke external tools. MCP, the protocol that connects AI to external tools and data sources, also played a role in coordinating those tools. For small businesses without a professional production team, its value lies not in “AI can edit videos,” but in reducing the manual transfer of work between multiple software applications.
Industry view
We believe this type of tool changes the entry point to work first. Previously, people opened separate drawing, voiceover, and editing applications. Now, AI can orchestrate the workflow. Connecting asset generation, subtitle checks, and script repair into one pipeline can genuinely reduce repetitive labor and accelerate content experimentation for small teams.
The risks are equally clear. The test author could not play the final video smoothly, and the automated system still required a human to download the file and validate it with a local player. Automatic environment installation can also introduce permission, security, and dependency-conflict issues. In other words, a completed workflow does not mean the result is deliverable. Errors may simply have moved from the chat window into files, scripts, and system environments.
Moreover, this test included only one case and one 3D cartoon style, which is not enough to prove that the tool can reliably meet real commercial requirements. Copyright provenance, character consistency, platform review, and cost accounting also did not receive sufficient attention in this test.
Impact on regular people
For enterprise IT: AI agents are beginning to modify local environments and execute code. Enterprises need to reassess permission isolation, software supply chains, and operational logs rather than treating these agents as ordinary productivity software.
For workplace professionals: Scriptwriting, voiceover, subtitling, and simple editing can be compressed into a single task, but final review and business judgment still require human accountability. The value of work will increasingly center on defining goals and enforcing quality.
For the consumer market: Businesses can test short-form video at lower cost, but “automatically generated” will rapidly increase content supply. On platforms, brand distinctiveness and credibility may become scarcer resources than output volume.