Learning From Architecture

I’m a massive fan of architecture. You get the chance to reshape the world and create places for human community to develop.

Now, instead of founding an architecture firm, I went ahead and created a non-profit research lab. Despite the many differences, I’ve recently been curious about the parallels between the two: They’re project-based, you receive awards for the best works, and you’re dependent on competing for funding to be profitable.

Similarly, the impact / brand of both firms is highly dependent on a few striking projects, such as BIG’s Copenhill and Apart’s DarkBench. Let’s first understand how a firm like BIG works:

Case: Bjarke Ingels Group (BIG)

The culture and philosophy of BIG is called ‘Yes is More’, a principle of experimentation, positivity, and enhancing latent potential. Their ‘pragmatic utopia’ idea means each project encapsulates a single inspiring idea, often in connection with the land, exceptionally executed.

To support this vision as they’ve scaled, they added new departments to include new perspectives; BIG LEAP now consists of BIG Landscape, Engineering, Architecture, and Planning and Products. Their focus is on fantastic outcomes that are great to live in, not just beautiful designs (sorry, architects).

These perspectives also extend to a diverse workforce with majority women leadership and strong benefits, encouraging novel and inclusive designs. This includes BIG’s CEO Sheela Søgaard, a former McKinsey consultant and CEO of another creative firm. Today, Ingels has the position of Creative Director and Founding Partner while Søgaard focuses on growth and financial leadership¹.

The growth they’ve experienced in the past 19 years² as a result (~$220M annual revenue in 2024) has helped them win projects all over the world. They maintain a global perspective on architecture while maintaining a deep respect to local ideas, respecting tradition and successfully integrating with the local communities.

When it comes to organization structure, you’ll notice that Bjarke Ingels is credited as the partner in charge on nearly every big project on BIG.dk. This is in collaboration with other ‘senior authors’, project managers, and supporting staff. Ingels doesn’t have an office in their HQ and mostly moves around to work with different teams. It’s a hierarchical structure where there’s no doubt that Ingels is the creative leader and senior partners are in charge[ˆ5], often promoted from within.

[ˆ5]: Since they’re named ‘senior partners’, they probably receive equity as a result of being the most competent individuals at the firm. This is also how partnerships work in funds.

When it comes to presenting the results of their work, BIG’s ideas are communicated incredibly simply. No complex words or artsy gate-keeping, just beautiful visuals, diagrams, and renderings. This stands in contrast to much of architecture³ and their functionalism and clear presentation makes them more attractive to clients.

While they stay clear, BIG keeps their originality, both going beyond and violating boundaries of the project specs in many competitions, winning as a result. They also have an in-house R&D unit for radical designs called BIG Ideas.

As a firm, architecture companies work on a project-to-project basis, earning money through competitive bidding on project outlines, fee-for-service, and from winning requests for proposals (RFPs). Success is measured by client satisfaction, profit, and awards.

Architecture firms work with long implementation timelines (sometimes decades), large average contract sizes (such as 5%-20% of construction costs), and clear deliverables (floor plans, renderings, and engineering simulations).

For research labs, a project most often take maximum one year (with citations showing impact over time), the contract sizes are usually below $100k (though grants can be in the millions and last 5 years), and it’s often difficult to predict the complexity of deliverables (most often in the form of peer-reviewed papers).

However, let’s take a look at a real example:

Case study: Apart Research

Apart follows a few core principles (without having written a whole book about it): Optimism, meritocracy, and humans-first. The mission of Apart is to be the fastest way to impact for aspiring AI safety researchers anywhere.

As a result, there’s a unique culture of both academic precision and product thinking (or, ‘hacker culture’), which invites a holistic perspective on most problems that Apart faces as an organization.

Instead of single senior partners responsible and credited on every project, each paper is community-led and run by a lean team of researchers with one designated lead author. Apart’s staff are often last authors (which indicates senior involvement in academia).

Apart applies the principle of ‘simple idea leadership’ by hosting research hackathons: In collaboration with topic experts, an event is defined and hosted with a keynote talk and two days of global hacking. Each team ends up with a pilot research report (example) reviewed by senior researchers in the field.

Through the sprints, Apart invites diverse perspectives on global topics in AI safety. After the reward ceremony, the best teams are invited to publish their work in academic venues guided by the Apart team.

As a non-profit AI safety research lab, Apart’s revenue is based on philanthropic donations from sponsorships, requests for proposals, and donor relationships. Our success is measured in quality-adjusted research impact, community development, and awards.

After three years, Apart has grown into a 8-person remote-first lab with 30 researchers working on projects at any one time and thousands collaborating during research hackathons, leading to more than 15 peer-reviewed publications per year and 300 original hackathon research reports.

Now, what can a research lab like Apart learn from architecture firms like BIG?

Let’s go through it one by one:

Research: Besides ensuring that research is high quality and on time, ensure that the title, abstract, and main figures make the complex ideas in research as simple to understand as possible. Go above and beyond on the media associated with the most impactful projects. Maintain a strong visual and research identity. Create a great package, do as much as possible to make it impactful, and go on to the next project.

Culture: Stay completely focused on outcomes and expand strategically from there. Re-iterate the core principles and use them to create the research unique to your organization. Double down on everything that makes you special.

Leadership: Don’t let the principal investigator (PI) be the CEO. Hire a professional to lead growth, financial sustainability, and administration. Have a clear hierarchy of conceptual leadership and remove all administrative tasks for researchers while compensating skill and results adequately.

Financial sustainability: Establish a foundation for growth with financial boundaries around the research process - apply to all relevant RFPs, build long-term relationships with collaborators and funders, and don’t host an event or start a project if there’s no prospect for funding.

Creating tomorrow’s research lab

Research has developed a lot throughout the ages. In Renaissance Italy, you’d hide your results to ensure your patronage would continue in case you were challenged on your math skills. In ancient Greece, flexing was valid argumentation during Senate discussions.

During the Great Wars, research became a focused effort to create deadly weapons on a Manhattan Project scale, and in the late 20th century, grifters shared sensational research findings and went unpunished, while honest academics pushed science further than it had ever been.

Now, as AI reshapes research, the biggest funders of research are trillion-dollar companies, leading to less research freedom, lower publication rates, less open source, and more marketing.

As research changes once more, we are all responsible for staying 5 steps ahead. And we’ll make sure Apart does.

Related writing

• Go on a yearly hiatus
• Sentware
• My Tools