Blackmagic Does AI Differently
Most AI feature announcements in video production tools follow the same pattern: a demo with ideal footage, a capability described in marketing language, and a real-world implementation that works under specific conditions but fails in the situations you actually encounter on a production. I've tested enough of these to have developed a healthy skepticism about feature launches.
DaVinci Resolve 20, released in May 2025 with updates continuing through early 2026, takes a different approach. The AI features are embedded in specific, defined tasks within the editing workflow — not general-purpose generation, but targeted assistance for tasks that currently take editors significant time. Understanding which features are genuinely useful for production work, and under what conditions, requires going past the feature list.
AI IntelliScript: The Interview Edit Problem Solved
AI IntelliScript is the feature I find most compelling for the production work I do. Here is the problem it solves: you have a two-hour interview shoot with multiple cameras. You have a script or transcript of what you want the subject to say. Assembling a timeline where the right takes are in the right order, with alternatives available for review, previously required an editor to manually match the script to the transcribed audio and drag clips into sequence.
IntelliScript automates this. The AI matches your script text to the transcribed audio across all clips, identifies which takes contain the relevant content, builds a timeline with the best selected takes, and places alternative takes on additional tracks for the editor to review. The editor starts with an assembled sequence rather than a bin full of clips and a script document open on a second monitor.
For documentary-style commercial work, brand film interviews, testimonials, and any production that involves syncing written intent to recorded performances, this collapses the assembly step from hours to minutes. The editor's value shifts to reviewing the AI assembly and applying creative judgment rather than performing the mechanical matching task.
AI Multicam SmartSwitch: The Multi-Camera Cut
Multi-camera interview edits — three cameras on a subject, or two subjects in conversation — have a standard mechanical step: deciding which angle to show when. SmartSwitch handles this mechanically. It analyzes audio and lip movement to identify the active speaker and automatically switches to the camera angle showing that speaker.
The output is not a finished edit. It is a first pass where the camera selection is handled and the editor can focus on pacing, rhythm, and moment-to-moment choices rather than sorting through angle options for each line of dialogue. For content that benefits from interview-style presentation — brand stories, customer testimonials, thought leadership pieces — the first pass assembly time drops significantly.
The edge cases are where human judgment still applies: overlapping dialogue, pauses where the non-speaking subject's reaction matters more, moments where the technical best angle is not the editorially correct one. SmartSwitch gets you to the starting point quickly. The editor makes the final decisions.
AI Audio Assistant and Dialogue Matcher: The Mix Problem
Audio in commercial production consistently produces the same problem: footage shot on different days, in different environments, on different microphones, with different room acoustics, needs to sound like a coherent whole. Matching audio levels, dialogue tone, and room environment across clips is a time-intensive step that often gets compressed when production schedules tighten, resulting in audio that betrays the edit to any listener paying attention.
The AI Audio Assistant analyzes the full project's audio and builds a professional mix: organizing tracks, evening dialogue levels, adjusting music and sound effects relative to dialogue, and producing a mastered final mix. The AI Dialogue Matcher goes further — it adjusts tone, level, and room environment of dialogue clips to match each other, even when recorded in different conditions on different shoots.
For short-form commercial work where audio post is a single step in a fast turnaround, these features reduce a multi-hour task to something that can be reviewed rather than built from scratch. For content heading to broadcast where a dedicated mixer is part of the workflow, the AI pass provides a quality baseline that the mixer works from rather than starting from zero.
The AI Music Extender addresses the timing problem every editor encounters: you have a music track that is 15 seconds shorter than the video, or 30 seconds too long. Music Extender analyzes the track and generates four extension or shortened versions that match the original's style and structure. You pick the version that fits. The days of looping a section awkwardly to fill time are essentially over for the audio sources the model handles well.
AI Animated Subtitles: The Caption Upgrade
Subtitles on social video went from optional to expected in the past three years. The AI Animated Subtitles feature connects Resolve's transcription engine to Fusion title templates — the subtitles animate as words are spoken, matching dialogue pacing. The implementation is the interesting part: you generate a subtitle track via transcription, then drag Animated Subtitle Fusion templates onto the track header to change appearance and animation style. The result looks intentionally designed rather than auto-generated.
For social content that will be viewed without audio — which is the majority of social video consumption — subtitles that are visually integrated with the content rather than legibility-only captions change the viewing experience measurably.
The Free Version Still Gets Most of This
DaVinci Resolve 20's AI features are available in the free version, with the exception of capabilities that specifically require the Neural Engine processing in the Studio version. The free tier gets IntelliScript, Audio Assistant, SmartSwitch, and Animated Subtitles. The Studio version adds depth map, advanced magic mask, and certain rendering acceleration features. For most production workflows, the free version covers the AI toolset that matters day to day.
Resolve 20 is currently at version 20.3.2, with ongoing updates through 2026 adding incremental improvements to the AI tools. If you're running an older version, the update is free and the AI features are immediately accessible after installation.
Sources: Larry Jordan — AI-Powered Features in DaVinci Resolve 20 | CineD — DaVinci Resolve 20 AI Features | Blackmagic Design — What's New in DaVinci Resolve
