Manifesto
How we think about AI.
A living document from Second Spring Design. Last updated April 2026.
We are not writing this from above the problem. We are building inside it.
We build Aventide, an AI platform for independent businesses.
We are small business owners building for small business owners, using the same tools, facing the same questions, and trying to make something more humane out of a technology that could easily go the other way.
AI can help people think, organize, create, and act. It can also dull judgment, flatten voice, concentrate power, invade privacy, and turn every saved hour into more work.
This page is where we say plainly what we believe AI should and should not do — and how those beliefs shape what we build.
If you have been quietly skeptical of AI — not because you hate technology, but because something about the current direction feels wrong — this page is for you.
Our belief is simple:
“AI is going to reshape work either way.
The question is where the power goes.”
It can flow upward: to fewer platforms, fewer employers, more automation, more dependency, more sameness.
Or it can move outward: to independent businesses, small teams, local communities, and people building work they actually own.
Second Spring Design exists to help make the second path more possible.
These are our four commitments.
In this document
1.
Room
AI should give you room, not make you run faster.
Almost every AI company is selling speed.
Do more. Ship more. Produce more. Automate everything.
We think that is the wrong pitch for the people we serve.
Most independent business owners did not go independent so they could squeeze more output into every hour. They did it because they wanted the work to mean something. They wanted more control over their time, their relationships, their craft, and the life around the business.
The right job for AI is not to fill every freed-up hour with more work. It is to give some of those hours back.
AI should help with the work that quietly drains people: repeated explanations, blank-page drafting, scattered context, admin follow-up, planning friction, and the dozen tools that never quite talk to each other.
We measure good automation by what it lets you stop doing, not just what it produces.
Our commitment
We build AI to create breathing room, not a faster treadmill.
2.
Agency
AI should make you sharper, not smaller.
The most important thing about a business owner is their judgment.
The version of AI that worries us is not only the dramatic one where machines replace people. It is the quieter one — where people slowly stop deciding, stop noticing, stop questioning, and stop trusting their own read on a situation because the model always has an answer ready.
We design our tools to take the drag off decision-making, not the responsibility of deciding.
That means AI should ask questions back, surface tradeoffs, show options, and help turn unclear thinking into usable next steps. When AI produces something, the owner should still be able to review it, challenge it, edit it, and decide whether it is right.
This is one of the hardest lines for us, because our product does execute work. It creates plans, drafts documents, queues workflows, and helps move things forward. That is useful. It is also exactly why we have to be careful.
We will not always get the line right between useful execution and quiet erosion of judgment. When we get it wrong, we want to hear about it, and we want to fix it.
Our commitment
We automate the drag around decision-making, not the responsibility of deciding.
3.
Grounding
AI should be grounded in your reality, not averaged from everyone else’s.
A lot of AI feels wrong because it sounds confident without being connected to anything real.
It gives polished answers without knowing the business. It writes in a voice that belongs to no one. It turns uncertainty into paragraphs. It makes every company sound like the same company.
That is not what small businesses need from AI.
Aventide is built around actual business context: facts, goals, customers, documents, constraints, offers, plans, and history.
The goal is not to make a business sound bigger or more generic. The goal is to help people see and express the real shape of their business more clearly.
That also means being clear about limits.
If the system does not have enough context, it should say so. If an output is based on assumptions, those assumptions should be visible. If something needs human review, it should not pretend otherwise.
Some Aventide features rely on third-party AI providers. That means parts of user interactions may be processed by those providers so the product can work. We will not pretend otherwise.
We also use anonymized and aggregated patterns across businesses to improve recommendations over time. For example, the system may learn what strategies, channels, pricing models, or workflows tend to work for certain types of businesses. The goal is to learn from patterns without exposing individual businesses.
Business context should remain with the business. Users should be able to correct it, export it, and leave with it.
Our commitment
Aventide should earn trust by staying grounded — in real context, visible assumptions, and limits people can understand.
4.
Independence
AI should make small businesses harder to crush, not easier to replace.
The last twenty years of technology have often made independent businesses more dependent.
Dependent on search algorithms. Social platforms. Marketplaces. Ad networks. App stores. Delivery platforms. Payment systems. Software stacks they do not control.
AI could make that worse.
It could become another layer of dependency where a few companies own the models, the infrastructure, the customer relationships, and the economic upside.
But it could also run the other way — if people deliberately build it that way.
A solo operator or small team can now access forms of leverage that used to require a much larger company: strategy support, marketing help, financial modeling, planning, customer insight, content production, operations, and follow-up.
That matters for more than productivity.
When independent businesses thrive, more money stays closer to the people and places they serve. Skills stay distributed. Communities keep more character. More people get to build work they own instead of renting their future from a platform.
The point is not to make every small business act like a giant corporation. The point is to give small businesses enough leverage to stay independent, specific, and alive.
We do not want to lock users in. Data should be exportable. Context should belong to the business. If a better tool comes along, it should be easy to leave.
A tool that helps people keep more power has to behave that way itself.
Our commitment
The better AI future is not five platforms serving everyone. It is thousands of independent businesses serving their people well.
Restraint
AI is not weightless.
It uses energy. It uses water. It depends on data centers, chips, supply chains, and infrastructure that most users never see. Any serious way of thinking about AI has to admit that.
We do not have a perfect answer to this yet.
But we do have a principle:
“Use the smallest sufficient intelligence.”
Not every task needs a frontier model. Not every workflow needs automation. Not every saved hour should become more output. Not every business problem should become another stream of AI-generated content.
We want AI to be used where it creates real leverage: clearer decisions, less repeated work, better planning, more useful execution, and fewer wasted cycles.
We do not want to use AI just because we can.
The environmental question is not separate from the human question. A tool that floods the world with more noise, more spam, more campaigns, more dashboards, and more fake productivity is wasteful in more than one way.
Good AI should reduce waste — wasted time, wasted attention, wasted effort, wasted motion — without pretending the technology itself has no cost.
This is an area we will keep learning about and improving.
Our commitment
We will use AI deliberately, not automatically.
The product tests
These commitments only matter if they shape what we build.
So these are the questions we want to ask before shipping:
Room
Does this give time back, or create more work?
Agency
Does this sharpen judgment, or replace it?
Grounding
Is this rooted in the user’s real context, or generic output?
Independence
Does this reduce dependency, or create it?
Restraint
Does this use AI because it is useful, or because it is available?
We will not always get these right. But these are the tests.
What we have changed our minds about
This section will grow over time. When our position on something shifts, we will log it here, with a date and a short explanation of why.
Empty for now — this is a new document.
Why we wrote this
We wrote this because the AI conversation often feels split in two.
Too often, people build as if the risks are imaginary, or critique as if the tools are useless.
We do not think either posture is enough.
The work we care about is in the middle: building with AI while taking its costs seriously. Using the technology, but designing against the parts that make people smaller, businesses weaker, and platforms stronger.
If this is the kind of AI future you want to help build, we would like to hear from you.
— The Second Spring Design team
Second Spring Design
Human-scale AI for independent businesses.
Company
Product
Legal
Privacy
Terms
Contact
© 2026 Second Spring Design. Built for human-scale businesses.