Make with Notion 2024: Shaping our work with AI (Dan Shipper, Sarah Guo, Daniela Amodei)

AI is reshaping creativity less by replacing human imagination and more by removing the “overhead” that drains attention—while also expanding who gets to participate in creative work. Across the conversation, Claude is positioned as a general-purpose assistant that handles administrative and tedious tasks (email triage, proposal review, drafting outreach), freeing people to focus on the interpersonal and aesthetic parts of their jobs. The practical payoff is immediate: creators spend less time on drudgery and more time on the work they actually care about.

That shift also changes who can create. Sarah Guo argues that creativity is widespread—everyone has taste and a point of view—but technical skill often blocks people from turning ideas into finished outputs. As AI tools lower the barrier to execution, more people can express ideas in writing, software, visual arts, and video without needing years of specialized training. A recurring example is scaffolding: someone who can’t code can still mock up a basic website, or someone stuck in a draft can use Claude as a “thought partner” to unstick the next step.

Dan Shipper and Daniela Amodei add a second layer: AI doesn’t just accelerate tasks; it can function like a companion in the creative process. Writers use Claude to get unstuck when they don’t know where a story should go next. Designers and other creatives use it to explore tone, recommend directions, and even help articulate what they like—turning vague taste into words that can guide future work. Amodei describes Claude as both a coach and an interactive partner, including for “unimpressive” personal creative pursuits, and notes that users often report a sense of companionship rather than mere utility.

The discussion then widens to how creative tools evolve. Guo highlights a “capability-first” pattern: builders start with powerful capabilities even when the user-facing experience can’t yet meet every expectation. As tools become easier and cheaper—sometimes dramatically—people stop trying to recreate the old “ideal” outputs and instead create what the new tooling makes feasible. That feedback loop pushes product teams to iterate toward more controllable, editor-like experiences, including regeneration and fine-grained editing.

Personality and interface design emerge as a key “ergonomics” theme. Amodei explains Claude’s approachable, quirky tone as a deliberate product choice: it’s meant to feel usable every day, like a comfortable tool, and to adapt to user context and examples. Guo points to the popularity of AI companions as evidence that humans respond to natural language and social interaction cues. Still, the group draws a line between friendly assistance and problematic anthropomorphism, arguing that the right level of personality depends on use case—customer service, therapy, or creative work all demand different boundaries.

Finally, the panel frames adoption as a learning curve that keeps shrinking. As models improve and developer tools become more capable, even people who once felt “this isn’t for me” find new ways to use them. The shared conclusion: the biggest winners are those who stay open-minded, treat AI as a partner in practice, and use it to translate taste, context, and intent into outputs they couldn’t produce before.