LO
Content creator — tasks, one by one
The unit of analysis is the task, not the job title. Each one below carries its direction, whether the judgement rests on evidence or on platform inference, the reasoning, and what it does not establish.
Every task on this page#
Finding the idea the audience will watch
Being augmented✓ Evidence-backedReading the comments, the trends and the numbers, and deciding what to make next that this audience will click on and stay for.
Idea tools are built into the platforms themselves and brainstorming is one of the most common uses creators report for generative AI. But the platform that offers the ideas also refuses to pay for content made from generic templates without the creator's own insight or perspective — the idea has to be theirs for the channel to earn.
The survey figures are self-reported and one comes from a tool vendor; nothing here measures how many ideas now start with AI or whether they perform better.
Writing the script
Being augmented✓ Evidence-backedTurning the idea into what will be said: structure, hook, jokes, the line that makes people stay past the first minute.
Chatbots are the tool creators most often name, and YouTube explicitly allows using AI to edit scripts and does not require disclosure for AI used in scripting. What it does not pay for is content that gives the impression of mass production without the creator's original, authentic insight — so AI can draft, but the voice that makes a channel worth paying for has to be the creator's.
Platform policy decides what earns, not what works; nothing here measures how much of creators' scripts are now machine-drafted.
Being the face and the voice
Still human-led✓ Evidence-backedPresenting on camera or on the mic as a person the audience recognises and trusts — the reason they subscribed.
Synthetic presenters and cloned voices exist, and the rules now single them out: YouTube requires creators to disclose realistic altered or synthetic content such as a generated voice narrating a video, the EU requires deployers to disclose deep fakes, and China requires users who publish AI-generated content to label it. A channel built on a real person's face and voice is built on the one thing the rules make a machine declare itself.
Disclosure is not prohibition: labelled synthetic presenters are allowed, and nothing here measures whether audiences prefer real people or how many channels use synthetic faces.
Editing, thumbnails and captions
Being augmented✓ Evidence-backedCutting the footage, adding graphics and sound, making the thumbnail and captions — often the most hours per video.
Editing and enhancement are among the most common uses creators report for generative AI, and automatic captions are standard on the platforms. The creator still decides the cut, the pacing and the thumbnail that decides whether anyone clicks.
The usage figures are self-reported by creators, one survey is from a tool vendor, and neither measures editing hours saved.
Dubbing and translating for other languages
Automating≈ Platform inferenceMaking the same video watchable in other languages — subtitles, voice-over, a second channel.
This is the part the platform now does itself: YouTube says automatic dubbing is available to everyone in 27 languages, and that in December 2025 more than six million viewers a day watched at least ten minutes of auto-dubbed content. What used to require hiring a translator and voice actor, or running a second channel, is becoming a setting.
The figures are the platform describing its own product; they count viewers, not how many creators rely on it or whether its quality holds for humour and tone.
The community
Still human-led✓ Evidence-backedReplying to comments, running the live chat, moderating, and being someone the audience feels they know.
Moderation filters and reply suggestions help, but the value of a community is that the audience believes the person is there. US law now treats the faked version of an audience as deception: buying fake followers, views or likes, including those generated by bots, is a violation when the buyer knew or should have known they were fake.
The rule is about faking the audience, not about how creators answer it; nothing here measures how much community work is automated.
Sponsors, disclosure and platform rules
Still human-led✓ Evidence-backedFinding and negotiating sponsorships, declaring them properly, and keeping the channel inside each platform's changing rules.
The rules that decide whether a channel earns are written for a person who answers for them: in the US a creator's material connection to a brand must be disclosed clearly and conspicuously, and endorsers can be liable for failing to disclose; platforms decide what is eligible for money and change it. Keeping inside those rules, and negotiating with the brands, is work that grows as the rules multiply.
The endorsement guides are guidance on how US law applies; nothing here measures how much time creators spend on sponsorships or compliance.