Generative AI is evolving rapidly. This makes it challenging for organisations to work out where to apply these new technologies in the short-term, and for their developers who need to continuously upskill as new iterations of generative AI models, frameworks and products are released.
One of the ways in which organisations can quickly understand new waves of generative AI is through rapid prototyping. This involves quickly building working versions of products or tools that incorporate generative AI in some way, and allows organisations to:
This rapid prototyping can be considered as the very early stages of more extensive agile product development: it is primarily about fast learning and reducing technological and product uncertainty, which in turn will enable quicker, more focused product development.
Here in Deloitte, we have several teams that focus on rapid prototyping, both to ensure Deloitte remains at the forefront of new technologies, and to help our clients do the same.
One of these teams is X Lab, which is part of Deloitte Digital.
The X Lab team have been experimenting with AI for over a decade, including more recently with generative AI. We’ve been rapidly building generative AI prototypes both internally within Deloitte and externally for our clients, and in doing so, developed a deeper and more sophisticated understanding of the underlying technologies, and how they can deliver value within complex commercial organisations.
One example of an early internal prototype that we developed was an image generation plug-in within Miro. For those unfamiliar with Miro, it’s an online whiteboarding tool, and one that’s used extensively in Deloitte for collaborative brainstorming and idea generation.
Could we integrate cutting-edge AI image generation into Miro so that our Deloitte colleagues could access this technology within their existing workflows?
This would allow users to generate images within a single application, rather than having to combine Miro with other browser-based tools. And whilst this would only provide a modest uplift in productivity compared to existing available products, it represented a quick and easy initial starting point to experiment with new image generation models.
We started with two workstreams in parallel:
We quickly found that there were multiple AI models that we could use, including both proprietary and open-source models. We settled on an open-source model, as it provided sufficiently high performance with a licence that permits commercial applications.
We then designed and built a simple user interface that we connected to a Python backend to host the AI model. A key part of the development process was also building an architecture to host and serve the app securely and reliably, especially to handle any errors, and to protect credentials like API keys.
After successful initial development and testing, we released the app to 32 separate teams across Deloitte.
In the weeks following its release, we followed up with teams to learn more about how they were using the app. Some were simply playing around with the tool as shown in Figure 1. Some used the tool to generate original images for their work as shown in Figure 2. And some even discovered a new artistic passion, producing the result shown in Figure 3.
Figure 1. Prompt: Magical Penguin
Figure 2. Prompt: Logo for a moodmeter app
Figure 3. Prompt: masterpiece, best quality, composition of human skulls, animals’ skulls, bones, ribcage, jellyfish orchids and betta fish, (orange and blue bioluminescent), intricate artwork by Tooth Wu and wlop and beeple. octane render, trending on artstation, greg rutkowski very coherent symmetrical artwork, cinematic, hyper realism, high detail, octane render, symmetrical, fan art, dramatic lighting, volumetric lighting, highly detailed, wind:1.2, 4k, 8k
This kind of app was almost unimaginable a few years ago, but now it is relatively quick and easy to produce with the rapid advancement of image generation tools, including their availability as open-source models and their ease of integration into products. This will continue to contribute a huge amount of value in society by making image generation available to almost anyone.
This experiment also demonstrated that the value of generative AI is not only in the models themselves, but also in the products that integrate these models into existing workflows. These products are essential in unlocking the value of generative AI for everyday users, and effective product development will be an essential part of realising the potential value of on-going advances in generative AI.
For our team specifically, this prototyping provided an opportunity to quickly experiment with an ever-increasing range of generative AI models. It helped highlight some of the unknown unknowns in generative AI product development, which – for example - will enable us to estimate more reliably the amount of time and effort needed to incorporate AI image generation into future products.
We have since developed several larger and more complex prototypes that incorporate generative AI, including as part of an interactive Paddle experience that we developed with the Lawn Tennis Associate (LTA), and which will feature in its own forthcoming blog post in a couple of weeks.
If you would like to learn more about rapid prototyping and how we can use it to accelerate your generative AI journey, please do get in touch.