AI’s threat to knowledge work has emerged faster than expected. Agencies and studios with teams of copywriters, coders, designers, strategists, and researchers now face an ultimatum that is no longer optional: Embrace and thrive or ignore and sink.
Any work that can be clearly articulated is now promptable and replaceable.
DO NOT think your clients are oblivious; these questions will be coming (if they haven’t already):
Are you guys using AI? How can we use AI? Is your work getting quicker now? Do we need a copywriter anymore? Can’t you just do that with AI now?
How you respond as a studio owner, creative, or developer will dramatically shape your client relationships and bottom line over the next 1-2 years.
Simply adding AI won’t do the trick; what will is how you bring your clients on the journey. But first, some context.
Meet Al
Lowercase “L”, not a capital “i”. Think Alpuccino.
Al is the self-improving new kid on the block who is basically fucking with the program on every conceivable level.
2020-2024: “Let me come up with a basic marketing strategy”. Harmless, intriguing.
2024-2025: “I can review and fix that code, no sweat. Review and accept the changes”. Useful, novel.
Al clocked off for xmas, for some family time, like always… he was chill…
Feb 2026: “While you were sleeping, I made a few SaaS apps to help your business, created pull requests for your Jira bugs, autonomously traded on polymarket, organised your finances, created video animations for your social media, drafted email sequences and booked an anniversary dinner”
Al went from a harmless Frank Grimes to an unhinged Akira-like beast in 2 months. He’s now working 24/7, munching tokens, performing tasks, passively accelerating life, modifying the fabric of reality.
The human is being squeezed out of the loop, and it's Al's journey we're now strapping in for.
Grief cope
We’re now simmering in the bargaining phase of the standard grief processing timeline, an inflexion point where we ask “what if”, attempt to regain control, and make mental deals in an attempt to change the outcome.
Minor cracks are appearing at a time that will go down as a potential emergent moment for AGI.
So where are the cracks?
The value problem: If everyone adopts similar tooling, the outputs become similar, the expectations align, and soon, pricing is driven down.
The slop problem: Al produces work that passes the glance test but crumbles under scrutiny. Copy that sounds confident but says nothing. Code that runs but breaks when used in a browser.
The tech problem: Understanding and fully adopting AI is technically challenging for most people.
The volume problem: Autonomy means more output at scale. More output means more reviewing and steering. Quality and taste require humans. Humans have become the blocker.
An example of the volume problem.
OpenClaw is roughly 3 months old with 4,200 pull requests.
NextJS is over 5 years old with only 1,300 pull requests.
The majority of these pull requests (code changes) are now submitted by AI agents, not humans.
This is a glimpse into what autonomy means and the flow-on it has. We’re not even at mass adoption yet. Normies (real people) have yet to experience the cracks; they’re still watching from the sidelines.
The new opportunity for agencies and studios is crafting the experience of integrating AI into clients’ ecosystems.
Getting real
Currently, agencies and studios are rapidly adopting AI across their workflows. It’s a subtle, internal-only thing for now, but will soon become front and centre. This will lead to a trend of underbidding for work, driving down the collective value.
How AI is layered in will become the differentiator.
Simply having AI do things won’t be a competitive edge. In a growing landscape of flakes, charlatans, and AI-fueled charlatans, there are big opportunities for folks who understand service in the truest sense.
Also, a stern reminder: if you’re just flogging the AI horse but the work isn’t objectively better, it won't fly.
The biggest value in service will literally be the actual service. Funny that. So what does it look like?
Experience in the Loop
Manpower (resources) and industry intel (knowledge) were previously limiting factors for agencies and studios; this is now fading fast, and the playing field is flattening.
The new model elevates experience over deliverables. It’s about using AI to deepen the journey that will by proxy elevate the output.
Education: Helping clients understand what AI can and cannot do in their landscape. Become the guide, not just the vendor.
Co-creation: Collaborate to leverage AI on strategy, processes, ideation and mood-boarding.
Data building: Plan and create a shared source of truth that teams can use AI to fetch or create from.
Change management: Helping clients' teams adopt AI without fear, confusion, or chaos.
The cracks were purely opportunities in disguise. The agencies that figure this out early will own the next era. The rest will be competing on price with Al's clones.
Adoption of AI is no longer an if question, it’s a how question.
