Research and Development

Designers need Augmented Intelligence not Black Box AI

Sayjel Vijay Patel
January 15, 2021

An Industry in Dire Need of Disruption

The building construction industry faces an existential crisis. It is one of the least digitized industries and the world’s leading producer of C02 emissions. With rising temperatures around the globe and mass migration to urban centers, there is a dire need for a digital disruption that will enable more sustainable ways of working.

We as architects can learn from other fields that have made rapid strides in developing synergistic human-machine systems which exploit the positive aspects of human and AI-generated reasoning.

Over the past year, a number of AI tools have been developed that provide new insights into the environmental impact and performance of a design. Despite their great potential, several barriers to widespread adoption exist:

  1. Significant expertise is required to take advantage of their capabilities
  2. AI and other computational methods require narrowly defined design goals
  3. For clients — the ones financing the projects — the outputs of these tools remain opaque and not understandable.

At Digital Blue Foam (DBF) we develop new solutions to accelerate the world’s transition to better and more sustainable cities. We are designers and technologists with a strong sense of responsibility to drive a desperately needed revolution in architecture, engineering and construction (AEC) industries towards carbon-negative projects.

At the dawn of a new age of design, several fundamental questions remain:

  • How do we develop tools that give designers greater agency to promote sustainable designs?
  • What does the next generation of user interactions look like?
  • How do we connect the strengths of both human and machine intelligence in a design tool?

AI Makes No (Common) Sense

2019 was a prolific year for new Machine Learning (ML) applications . ML, the machine’s ability to infer outputs from structured, or, in the case of Deep Learning, unstructured data, has garnered massive hype and hysteria. This has led to the common misconception that if we collect and process enough data using the latest algorithms and computing power, intelligence will simply emerge. This is simply not the case.

Using a strategy of correlation rather than causation, ML identifies patterns incredibly fast, especially when the input data is well-structured. This is why it is so effective at playing video games like Flappy Bird. In most video games, rules and goals are consistent; unlike the real world, where the pattern will always stay the same. But what happens when the situation changes?

‘Causal’ intelligence is the basis of human smarts. While humans recognize patterns comparatively poorly, we excel at reapplying our past knowledge to new situations. So when Flappy Bird 2 gets released, we know it will have mostly the same rules and goals as the original, and we adapt quickly. Our AI, on the other hand, did not understand anything; it only followed a pattern. When the pattern is altered, the AI becomes useless.

Interfaces for Augmented Intelligence

Digital Blue Foam has developed a tool that can steer the AEC industry around a major bottleneck. Architects spend a lot of time learning and using inefficient drafting tools that actually distract them from their essential role of creating and experimenting with design options that cater to the plethora of needs of the site and project, be they social, environmental, or economic. Our tool uses augmented intelligence to free architects from this hassle and aid them in generating design, through the augmentation of the time-tested method of sketching.

We at DBF are major proponents of augmented intelligence’ as opposed to black-box AI. In this paradigm, AI becomes a tool to enhance human intelligence rather than replace it. While sophisticated AI systems are able to make decisions after analyzing patterns in ‘big data,’ they are only as good as the data humans give them.

Image for post

Augmented intelligence is already embedded within many well-known products. It is particularly evident in word-processing applications: MS Word’s spell-checker, the Hemmingway Editor, and Grammarly (an online grammar checker) are a few everyday examples of augmented intelligence. The ‘design ideas’ feature in Powerpoint, which uses augmented intelligence to help non-designers improve their layouts, is another example.

In our software, the user designs with the computer, using a pen interface. What ensues is something others have described as human-in-the-loop augmented intelligence with human-computer collaboration (Zheng, 2017).

I Human Design Approach

Image for post

The manual design approach stems from intuition, experience and logic. It is able to suggest solutions to complex problems, such as fitting different types of buildings in different parts of oddly-shaped plots.

II Computational Approach:

Image for post

A computational strategy has the ability to quickly generate solutions that follow a predefined logic. The nature of this approach is perfect for iterative tasks such as generating rule-based solutions and performing calculations and analysis.

III Augmented Intelligence (1+1 >2):

Image for post

An augmented intelligence (AI) approach combines the merits of the previous two strategies. It comprehends the designer intent, in this case through lines or nodes and applies rule-sets to generate highly- nuanced solutions

In less technical terms this means that as the designer sketches and the computer interprets the drawing as an input to constrain the generation of a new design. In turn, the designer observes what the computer generates. They can then sketch over top to edit the result, providing new input. By repeating this process, new solutions emerge which neither machine nor human could have imagined.

By applying the principle of augmented intelligence in our technology, we hope to connect common sense gained through a lifetime of human experiences with what machine intelligence does best: identifying helpful patterns in large data sets.


Steve Jobs was a huge advocate of augmented intelligence and this approach is responsible for much of the success of Apple products. In a 1994 Rolling Stone article, the visionary Jobs remarked: “Technology is nothing. What’s important is that you have a faith in people, that they’re basically good and smart, and if you give them tools, they’ll do wonderful things with them”.

Augmented intelligence is not a new concept but deserves more attention as we work out how to apply all of the amazing capabilities AI affords us. To address the existential crisis of the AEC industry, we must extend our definition of augmented intelligence and harness the true potential of AI: the sensitivity of designer intuition and the brute force of machine intelligence, combined.


  • Zheng et al (2017) ‘Hybrid-augmented intelligence: collaboration and cognition’. Frontiers of Information Technology & Electronic Engineering

Our latest news

Follow our journey as we accelerate the world's transformation to better cities.

Better Design Awaits

Get started for free and find out what you can accomplish with the power of
Digital Blue Foam at your fingertips
Try Now