A night on the jobsite

It's 6:30 PM. Julien, a construction project manager, stares at his screen — Word open — feeling that familiar sensation: exhaustion before even starting. Plumber photos scattered across WhatsApp—47 images to integrate. Handwritten notes in his field notebook. An email from the electrician buried in Outlook. 12-pages that need re-formatting, and a checklist to copy-paste from OneNote.

Time itself isn’t the weight—he knows it'll take about 2 hours. It's the activation energy. That huge psychological friction to start, assemble, and structure everything. He sighs. Again.

Tomorrow, his stakeholders will jump straight to the PDF conclusion... then close it immediately. The same defects will reappear next week.

This scene repeats in thousands of companies every day. There are tasks that take 20 minutes... but stay on our to-do lists for 5 weeks. We tell ourselves we lack time. But often, we lack momentum.

What if this friction disappeared? What if those 8 avoided hours became 30 fluid minutes, and reports were finally read and transformed into action?

Voice AI has already bridged the gap — and it’s a revolution

Voice notes spread through video meetings like wildfire. Speaker separation, automatic transcription, synthesis with key points, automatic integration into CRMs and ERPs—this is what transformed videoconferencing.

But the real change isn't in transcription speed. It's in eliminating the friction to start documenting. Before: enormous energy to take notes then transcribe them properly. Now: zero psychological resistance to capture—AI does the rest.

As Sam Altman, CEO of OpenAI, observes: "Already we live with incredible digital intelligence, and after some initial shock, most of us are pretty used to it. Very quickly we go from being amazed that AI can generate a beautifully-written paragraph to wondering when it can generate a beautifully-written novel [...] This is how the singularity goes: wonders become routine, and then table stakes."³

Voice AI has crossed this threshold: it became so light that it disappears. No more energy needed to start—you speak, it structures.

"On a construction site I can take 50-100 photos in two hours; then I only send myself the ones that will really serve the report." — QAQC Manager, Construction

Imagine the hours saved... but more importantly, imagine the number of reports that now exist because the entry barrier disappeared.

The real question: What are we avoiding?

In theory, AI makes capture trivial: automatic transcription, image recognition, pre-synthesis. In practice, three obstacles transform the promise into a mountain and bring us back to avoidance:

The field amplifies friction

In construction, this friction takes on a particular dimension. Unlike office work where you control your environment, the field imposes its constraints: weather conditions, safety emergencies, crew rotations, multiple stakeholders.

On a construction site, when a problem appears, it must be addressed immediately—not in 3 weeks when you've finally filled out the report. This timing creates an enormous gap between the pace of reality and that of documentation tools.

"In the morning we do the site walk, take 50 photos. In the evening we look at 5. The report comes out 15 days later, the problem is already solved or has gotten worse." — Project Manager, Real Estate Development

This specificity explains why generalist solutions often fail in the field: they're designed for office workflows, not for the urgency and complexity of the physical world.

1 | The energy problem: reports no one opens

When data explosion masks inaction

As soon as you free humans from drudgery, the volume of data they produce explodes. This is exactly what happened with smartphones arriving on construction sites: give teams a camera, and you'll get an incredible volume of photos.

"Customer conversations multiply and fragment across email threads, support tickets, Slack messages. What was once a clear signal becomes buried in noise." — Linear⁸

At Finalcad, we had accumulated tens of millions of photos. We even published a construction observatory listing the most recurring problems by trade. An impressive wealth of data.

The bitter reality: repetition without learning

The most frustrating part? Observing that on each floor, the same errors repeated until the end of construction, without any learning curve emerging.

The problem wasn't lack of data—there were millions. The problem was the activation energy to transform this data into action.

As Paul Graham points out: "When I'm working on an essay, I spend far more time reading than writing. I'll reread some parts 50 or 100 times... And the easier the essay is to read, the easier it is to notice if something catches."¹

Teams avoided consulting reports because the mental effort to digest them exceeded their available capacity. The improvement cycle was broken: when documents are too heavy to read, even their author doesn't reread them.

"Documents aren't read, that's why we have a weekly 1.5-hour meeting and daily 30-minute check-ins to ensure we share what's done and to-do." — R&D Manager

The cruel reality: these findings are useless if they don't reach the right people at the right time, in a format that doesn't require energy to consume.

2 | The friction problem: software that exhausts

Nobody likes filling forms

"The formatting template restricts me: it imposes useless boxes, it takes longer than starting from a blank page." — Project Director, Real Estate

No human enjoys filling out form templates. Those rigid boxes, mandatory fields that never match field reality. Forms break the natural rhythm of thought. As Paul Graham explains: "The rhythm of good writing has to match the ideas in it, and ideas have all kinds of different shapes."¹

Field observations have different shapes—sometimes a photo suffices, sometimes you need a nuanced explanation. Forcing them into predefined boxes means losing the essence of what you really wanted to communicate.

Worse yet: forms prevent following your natural curiosity. Paul Graham explains: "The way I usually decide what to write about is by following curiosity. I notice something new and dig into it." On a construction site, the expert eye notices unexpected details—but the template forces them to find the right box first rather than dig into the observation.

"The best product skill consists of reading what isn't said; users describe symptoms, rarely the root of their problem." — Linear⁸

When you force collection into rigid formats, you increase activation energy to the point where users give up. They work around it, postpone, or fill it out mechanically (we've even been asked to pre-fill fields as "compliant").

And worst of all: nothing more boring than reading forms with scattered boxes on a page optimized for space rather than reading time.

Forms to fill out, completed templates to read—that's a job for AI agents, not humans.

Forced collaboration breeds resistance

We experienced this at Finalcad: inviting all project members to a single platform didn't create the expected productivity, even with free licenses.

Forced transparency increases psychological friction. Nobody wants to share a problem with the whole world rather than dealing with it directly with their colleague. Nobody wants to create a user account on imposed software to enter occasional information.

Slack's success proved this: by bringing security to users, they created more content on private channels than public ones. Exactly like WhatsApp—exchange works because it eliminates social and technical friction.

The hell of auto-generated PDFs that discourage reading

A document is made to be transmitted to someone for information, action, or decision. The problem with automatically generated documents? They require as much energy to read as they saved to create.

"Importing 20 photos into Word, numbering them, resizing them... that's what takes us the most time." — Commercial Director, Waterproofing

Inboxes get flooded with very long PDFs, with complex layouts that make quick reading impossible. PDF on smartphones—ubiquitous in construction—is a nightmare. You have to zoom, lose context, tables become unreadable.

Result: users end up not opening these regular emails. They prefer to wait for the next meeting. The writer ends up orally repeating what they took time to document.

"A document with only condensed text, I won't even have the courage to read it." — Project Manager

3 | The systemic problem: engagement vs. momentum

Momentum as hidden metric

Companies measure everything: time spent, number of reports, open rates. But they ignore the most important metric: momentum.

• How many important reports are never started?

• How many analyses remain in heads because the friction to document them is too high?

• What are your users avoiding doing?

These invisible forces create what Paul Graham calls a "gravitational field": *"It's more like a kind of gravitational field that writers work in. But every essayist works in it, whether they realize it or not."*² Current tools generate a field that unconsciously pushes toward avoidance. Users don't rationally decide to avoid—they experience this gravity.

Transformative AI doesn't just reduce time from 8 hours to 30 minutes. It reverses this gravitational field to transform avoided tasks into fluid actions. It's the difference between optimizing the existing and unlocking the impossible.

"25% of my variable compensation depends on how quickly I send the report, because it's what triggers the rest of the contract." — Energy Maintenance Manager

Seat-based pricing is crumbling

In the era of AI and IoT, raw data is no longer just human. Will we need to pay user seats for each AI agent?

With massive deployment of AI agents, many software solutions will be attacked on their fundamentals. "Systems of record" that collect data to structure and export it will be driven by AI agents. Their number of human users will decrease.

Mike Krieger, CPO of Anthropic: "The future of product metrics in AI isn't engagement — it's actual value delivered."⁵

Kyle Poyar confirms in his State of B2B Monetization 2025: "Seat-based pricing is down from 21% to 15%. Hybrid pricing is up from 27% to 41%."¹⁰ The reason? "The rise of AI results in value being decoupled from customers needing to buy more seats. In fact, if the AI works as intended, customers likely need fewer people and more AI."¹⁰

there.do: Flip the logic, restore flow

Based on this observation at Finalcad, we founded there.do with a conviction: separate free collection from assisted construction to eliminate all friction at the source.

Preserving tacit field knowledge

An experienced site manager "feels" when something's wrong—a concrete texture, an alignment that's off, a tense team atmosphere. This tacit knowledge, forged by experience, doesn't fit in any form. there.do allows capturing these intuitions in their natural form—a photo with a quick comment, a voice note in the stairwell—then AI helps structure these observations into actionable insights.

Humans need freedom to build.

Frictionless collection, assisted construction

At there.do, we decided to make collection completely free. Our mission is accomplished when activation energy to document becomes zero—and a document is created, shared, and read.

"Your tool is like a smarter Word. The notes integrated from mobile to web and the structured document formatting saves time." — Jonathan Le Paplous, General Contractor

Private collection by default to eliminate social friction

"What's good is that you make your report from your notes. My problem is that I actually have them everywhere." — Nicolas Foussier, Agency Director

This "private-first" approach doesn't just protect confidentiality—it eliminates psychological friction. No more activation energy needed to "phrase things well" from capture. You note freely, AI structures, and you share only when ready.

The natural rhythm of the field: On a construction site, information arrives in waves—20 photos at once, then an isolated remark, then an emergency. Traditional tools break this rhythm by imposing rigid structure. there.do respects the chaotic temporality of the field: you collect everything in the moment, show only the essential when wanted.

This approach frees speech in the field—more photos, more frank remarks, because there's no more self-censorship.

"I created a note on mobile, when I create my account on the web, everything is already there. That's really good. I feel like it's Notion, but simpler and faster." — Vedran Beric, Designer

Putting the reader at the center to trigger action

"Without a workbench, we risk using powerful tools in a vacuum." — Linear

Our vision: put the reader at the center, provide them an experience that requires no energy to efficiently consume a document. Fluid layout isn't cosmetic comfort—it's what allows good decisions to emerge.

Paul Graham reveals the hidden equation: "The impact of an essay is how much it changes readers' thinking multiplied by the importance of the topic."² Field reports deal with crucial subjects (safety, quality, deadlines) but have little impact because they don't change readers' thinking—too heavy to digest. there.do optimizes "change" by making reading fluid, thus maximizing real impact.

"The big advantage is tracked email sending, you know who read what. Between the diagnostician, the engineering firm, the owner, the architect and the contractors, everyone has different interests. And we must control who is involved in the project or who doesn't care at all." — Sébastien Dangin, Architect

• 20-minute report but only 5 minutes available? One-click summary.

• On the road? Listen to the document instead of reading it.

• On smartphone? Adapted layout without constant zooming.

• Who read what? Integrated reading tracking.

"It's super responsive. There's really this fluid interaction, this side of freedom to be able to move a lot of things. It's not heavy, it allows you to go fast. Because with the Office suite, I can't take it anymore. I have the impression of lightness." — Vedran Beric, Designer

This "reading moment" is currently ignored by most solutions. They excel at collecting and structuring, but neglect the final experience: that of the human who must make a decision based on this information without it costing them energy.

Agent systems to route information

"Linear will be the place where agents live and operate in the same loop as the team." — Karri Saarinen, Linear⁹

Traditional systems like Document Management, ERP, and CRM won't disappear, but will transform. These systems of record will remain essential, but will now be fed by AI agents rather than humans forced to fill out forms.

Field capture solutions also retain all their relevance for punctual moments of massive capture. They could offer direct export to there.do or see their PDFs automatically transformed into structured notes, ready to be selected to build documents that read effortlessly.

Choose your spiral

"Quality creates gravity: it attracts without needing to push." — Karri Saarinen, Linear⁶

Continue the spiral "more data → more pages → more friction → more avoidance"?

Or adopt a flow "frictionless capture → light report → immediate action"?

This transformation is already underway. Mike Krieger, CPO of Anthropic, reveals that "90% of Claude's code is now written by AI, and this has completely transformed how they build products. The bottlenecks have shifted from engineering (writing code) to decision-making (what to build)."⁵

AI becomes the friction reducer between capture efficiency and communication fluidity. Humans remain in their environment of creativity, while agents handle the administrative work in the background.

"It's clearly Notion, but not for geeks. There's more drag-and-drop with notes, you need less database creation. I think positioning as 'Notion less geeky' is better." — Baptiste Mullie, CEO

Conclusion: From friction to action — Reinvesting Attention

Mehdi Boudoukhane brilliantly theorized that "The Hidden Cost of Work Isn't Time. It's Energy."¹¹ At Finalcad, we lived this reality in the field: millions of photos captured, but the same errors repeating floor after floor.

Voice AI has crossed the threshold; action AI hasn't done it everywhere yet.

there.do wants to fill this void for all those who shape the physical world—engineers, architects, project managers—who know their real problem isn't time, it's the energy to transform their observations into actions.

"The time-consuming micro-tasks between taking notes and sending a report—gone." — Kay K.

This transformation doesn't happen by replacing the existing, but by creating bridges. Between field capture and decision-making. Between private note and shared document. Between intention and action.

"If the recipient remains silent for 48 hours, I send another email; if they still don't respond, I organize a video call to go through the report." — CSM, IT

As Dario Amodei, CEO of Anthropic, emphasizes: "People have a long-term important and curious advantage over AI: we are hard-wired to care about other people and what they think and do."

At there.do, we've made our choice. We build for those who shape the world, so the engineer focuses on solving technical problems rather than the effort to start or read a report, so data flows where it needs to, when it needs to, without friction.

"The value proposition is undeniable, not even a debate." — Frédéric

It's in this new era that there.do was born: to transform Julien's 8 avoided hours into 30 fluid minutes, and especially so those 30 minutes become so light they launch naturally.

Because a task that never starts is infinitely longer than a task that takes time.

The journey is just beginning. Step by step, note by note, we're bringing to life this new generation of documents that breathe the same fluidity as our thoughts.

there.do is in open beta. Transform your first report today at there.do.

References

¹ Paul Graham, "Good Writing", May 2025
² Paul Graham, "The Shape of the Essay Field", June 2025
³ Sam Altman, "The Gentle Singularity"
⁴ Dario Amodei, "Machines of Loving Grace"
⁵ Mike Krieger, Interview with Lenny Rachitsky, "Anthropic's CPO on what comes next"
⁶ Karri Saarinen, "Why is quality so rare?", Linear Blog
⁷ Karri Saarinen, "Design for the AI age", Linear Blog
⁸ Linear Team, "Building what customers need, not just what they ask for", Linear Blog
⁹ Karri Saarinen, "Building our way", Linear Blog
¹⁰ Kyle Poyar, "The state of B2B monetization in 2025", Growth Unhinged
¹¹ Mehdi Boudoukhane, "The Hidden Cost of Work Isn't Time. It's Energy.", The Quiet Force