AI creators.
Beyond the demo.

The question after “what can it do?”

Ethan Mollick’s One Useful Thing explores how AI changes work, learning, and the choices people make with it. The Wharton professor and Co-Intelligence author gives readers a way to think about the technology beyond a stream of product announcements.

For a brand, the useful lesson is to connect capability with behavior. What changes in a real task? What should the person still decide? His work is a reference for a thoughtful discussion, rather than a shortcut to a conventional endorsement.

Show enough of the work to make it useful.

Simon Willison’s writing connects new AI capabilities with experiments, code, and practical observations. The Datasette creator and Django co-creator gives technically curious readers a way to examine what a tool does and how it behaves.

This is a strong reference for a developer product with something concrete to demonstrate. An example repository, a reproducible task, or a carefully explained integration gives the audience more to work with than a claim about productivity.

Give people the explanation they keep coming back to.

Andrej Karpathy’s educational work moves between technical foundations and broad explanations of language models. His Deep Dive into LLMs and How I Use LLMs offer two useful entry points: understanding the system and seeing how someone experienced approaches it.

For a brand, this is a benchmark for depth and clarity. The creative opportunity is educational substance with lasting value. An association would need to serve that work; a conventional scripted product pitch would miss its appeal.

For the decisions a demo leaves out.

Chip Huyen writes about building machine-learning and AI systems, including the books Designing Machine Learning Systems and AI Engineering. Her work gives builders language for the choices around the model: data, evaluation, architecture, and the demands of an application.

That makes her a useful reference for infrastructure and engineering education. A strong campaign concept would start with a design problem that readers recognize and explore the options in enough detail to help them make a decision.

Make “better” something the team can examine.

Hamel Husain’s work focuses on practical AI engineering, with particular depth in evaluation and error analysis. His writing and teaching help teams look closely at the behavior of an application instead of relying on a few impressive examples.

For a developer-tool brand, the opportunity is to make a difficult engineering task easier to learn. A useful example should expose the decisions behind the evaluation, giving builders something they can adapt rather than a dashboard to admire.

Help someone become a builder.

Jeremy Howard co-founded fast.ai, whose practical courses make machine learning accessible through doing. The approach gives learners working examples early, then develops the understanding needed to go further.

That is valuable territory for tools that remove a real obstacle to learning or building. Compute, development environments, and educational products can have a clear role when the campaign helps someone complete an actual project.

Start with the task on someone’s desk.

Jeff Su’s work connects AI and productivity with the routines of professional life. His tutorials and resources focus on making tools useful in familiar situations, giving viewers a practical reason to try a workflow.

For a workplace product, the strongest brief is a task with a visible finish line. Show the input, the decisions, and the usable output. The viewer should be able to recognize their own work in the example.

Help the team move from interest to a useful first step.

Allie K. Miller’s work brings AI into business conversations through education, resources, and executive guidance. Her background includes AI roles at AWS and IBM, while her public work focuses on how organizations approach the technology.

That perspective suits a brand selling to teams rather than individual hobbyists. The campaign should make the adoption decision clearer: which task to start with, who needs to be involved, and what a useful outcome looks like.

Make the unfamiliar concept feel learnable.

Tina Huang combines an ex-Meta data-science background with content on AI, coding, careers, and self-study. Her explanations of agents and local AI models give viewers a compact way into technical subjects they may be encountering for the first time.

For an educational or developer product, the opportunity is a useful learning sequence. A clear concept, a small exercise, and a working result can make the product part of someone’s growing ability rather than another tool to collect.

Give a new tool a reason to be tried.

Matt Wolfe connects AI news with tool discovery through his videos, Future Tools, and The Next Wave podcast. That combination gives a new product several possible contexts: a demonstration, a comparison, or a discussion of where the category is going.

The strongest brief contains a task that can be shown clearly. A feature becomes interesting when viewers can see what it lets them make, how the process works, and where it fits among the tools they already know.

Put the announcement in a reader’s working day.

Rowan Cheung’s The Rundown is a creator-led publication covering AI news, tools, and practical learning. Its team format matters: the opportunity extends beyond one person’s social feed into a publication and educational resources.

For a launch, the useful question is what readers can do with the news. A concise introduction paired with a genuinely useful guide gives the product more substance than a feature announcement alone.

Meet the people deciding what to build with.

Shawn Wang, known as swyx, connects technical writing with the AI engineering community through Latent Space and AI Engineer. His work gives practitioners a place to discuss the tools, roles, and implementation choices emerging around AI.

This is a strong context for infrastructure and developer products that benefit from conversation. A technical demonstration, a workshop, or a candid discussion with the builders can give the product a place in the community’s working knowledge.