01
02
03
04Sections
AI has made creative production abundant. Diagnosis, judgement and decisions remain scarce, and that is where the value in brand and communications work now sits. A company that knows what it sells, to whom, on what evidence, gets more from AI than any company still working those questions out.
A founder can produce a competent landing page before lunch. A team can generate fifty headlines, three deck structures, a visual system, a research summary and a month of social posts without hiring five specialists. The marginal cost of plausible material has collapsed, and adoption has followed. Stanford's 2025 AI Index recorded a surge in business use of AI.
The hard questions now sit upstream. A company has to decide what it is selling, which audience matters first, what changed in the market, which claim it can defend, which evidence should lead and what to leave out. It also has to resolve its contradictions before anyone writes the page. AI is good at producing possibilities. Strategy is the work of refusing most of them.
Fluency has lost its value as a signal
For years, polished language carried an implicit signal of effort. A clear deck, a coherent website or a well-written article suggested that someone had spent time making choices. Generative systems weaken that signal. Fluent text, attractive layouts and competent summaries are all abundant, so a thing that sounds considered tells us less about whether anyone considered it.
That raises the value of evidence and provenance. A claim linked to a test, a contract, a dataset or an observed customer problem carries more weight because the language around it costs nothing. A specific point of view carries more weight because generic synthesis is cheap. A feature the company deliberately left out says more because endless feature lists take no effort. The credibility signal now is judgement the reader can see. We cover how to make that judgement visible in evidence-based storytelling.
AI amplifies whatever strategy is already there
A company with a clear position can use AI to deploy that position quickly. A company without one can use AI to produce ten contradictory versions of itself at speed. We think of brand strategy as an operating system for this reason. The organisation makes prior decisions about offer, audience, position, voice, evidence and boundaries, and AI then adapts those decisions into formats.
Without that source layer, each prompt starts a new company. The homepage becomes one version. The sales email becomes another. The investor deck leans into a third, and the founder's LinkedIn posts invent a fourth. Everyone looks productive while the market receives a moving target. AI exposes the inconsistency that was already inside the organisation. The decisions that form the source layer are set out in What Is Brand Strategy for Climate Tech?.
Build the deep source first, adapt second
The Story System uses a rule that grows more useful as generative tools improve. Build the deep source first. Adapt second. For a founder story, that means recording what actually happened before asking AI for a 200-word version. For a product story, it means fixing the mechanism, evidence, customer context and claim boundary before creating landing page copy. For a pitch, it means deciding what the investor needs to believe and which proof carries each step before generating slides.
AI should transform source material the company owns. When a prompt leaves a gap, the model fills it with something plausible, and that invention becomes the source by default. In climate and science-led work, a fluent invented detail can become a false technical claim very quickly. How one source layer serves investors, buyers, partners and hires is covered in How to Build a Story System.
Judgement starts with the constraint
One of the easiest mistakes is to ask AI to repair the visible artefact before anyone has diagnosed the problem. The deck is weak, so the team rewrites the deck. The website is unclear, so the team rewrites the homepage. The investor stopped responding, so the founder drafts a follow-up.
Sometimes that is the right intervention. Often the cause sits elsewhere. The product may not be competitive. The buyer may have no budget. The evidence may be thin, the project may not be financeable, the investor may be outside mandate, or the company may not have decided what it sells. We start diagnostic work with a stop rule for exactly this reason. Rule out the commercial constraint before diagnosing a communication failure.
AI makes that discipline more important because it produces a convincing communication repair in seconds. Speed makes the wrong intervention cheaper and leaves it just as wrong. The cases where a better brand cannot help are set out in When a Climate Company's Problem Sits Beyond Branding.
The moat is a chain of choices
People in creative work often look for one defensible advantage, whether taste, brand, proprietary data, distribution, community or expertise. For evidence-led communication, we see the advantage as a chain.
- Selection.Which problem deserves attention.
- Diagnosis.What is actually causing the problem.
- Evidence.Which facts deserve trust.
- Position.Which interpretation is useful and defensible.
- Sequence.What the audience needs first, second and third.
- Expression.How the idea becomes language, design or experience.
- Review.What to remove because it is weak, generic or untrue.
AI can support every link. The human value lies in owning the consequences of each choice.
Distinctiveness comes from material the model cannot infer
A language model learns patterns across enormous bodies of text, so it produces the centre of a category with ease. Ask for a climate tech homepage and the draft reaches for accelerate, unlock, transform, future, sustainable and scale. That is the statistically sensible answer to a generic prompt. Research points the same way. In an experiment published in Science Advances, writers given AI ideas produced stories rated as more creative, and those stories were more similar to one another.
Distinctiveness comes from material the model could not infer from the category alone. A founder's odd turning point. A customer contradiction. A proprietary result. A failed pilot. A decision the company made against category convention. A specific way the product behaves, or a market shift the team saw before others did. The better the raw material, the more useful AI becomes.
Once the truth is fixed, abundance has a productive role. AI can cut a reliable source story into website, investor, sales and hiring versions. It can test headline families, compress a technical explanation for different levels of expertise, find repetition across a 100-page site and flag unsupported claims for human review. The order is truth first, variation second. A film editor works the same way. Editing can build great meaning from footage, and a documentary editor still needs the footage to exist.
What this means for studios and for founders
Studios and agencies feel this shift directly. A client can now reproduce a plausible first draft of copy, a deck or a page internally. The durable value lies in helping the organisation make the decisions those outputs depend on. That means deciding which market story is real, which audience gets priority, how a platform becomes a product story, which evidence belongs in the first five slides, what the company can claim and how the brand holds up in five different rooms. Execution still matters. The economic centre has moved towards diagnosis, judgement and system design.
For founders, the most valuable prompt may be the one they never run. Before commissioning any piece, a team can test it against a few plain statements. A decision requires this material. The claim deserves the space. This audience needs another deck. The founder story belongs in public. If a statement fails, the piece waits.
When production was expensive, cost made some of these decisions automatically. Now that production is cheap, restraint has to be deliberate. Judgement is the moat because someone still has to decide which sentence deserves to exist, what truth it represents and what should happen once someone has read it.
Brighter Future helps climate and science companies make those decisions before the production starts. See our positioning work.
Sources and further reading
Tell us what needs to move.
Bring the brief if it is clear. If it is unclear, tell us where the work is stuck.
Bring us the problem