There has been a lot of recent confusion surrounding Improbable, the UK software unicorn backed by a16z and most recently, Softbank to the tune of over $500m. I’ve heard it characterised as a gaming company, an AI company, even a VR startup and consequently began trying to unpack the source of this confusion. Improbable is what it describes itself as, a company that provides software that helps create virtual worlds. The confusion seems to stem from a combination of the width of this description and the company’s trajectory and business objectives.
The width of the ‘creating virtual worlds’ is one of the key drivers of mystique around the company. It is a phrase that can be readily interpreted narrowly or widely, one narrow interpretation of the concept leads to viewing Improbable and gaming company based on seeing virtual worlds as game worlds. On the other hand the wide conception sees virtual worlds as just that, worlds that do not physically exist. Improbable really sits under this conception – the company’s software is suited for the creation of both game worlds and simulated environments in general. The width of the offering has, however, been artificially obscured by the company’s initial focus on the gaming sector as its anchor market. This initial focus is still illustrated by the fact that its product lines are split into ‘Spatial OS for games’ and ‘Spatial OS for business’.
This whole confusion led me to wonder why a tech company would position itself as a gaming company initially and subsequently expand into everything else. What, if anything, was unique about gaming that made it best suited for the rollout?
This post answers this question arguing that a games first go to market (like Improbable’s) is genius as the sector, by virtue of being the confluence of art and technology is the perfect incubator for technological innovations with wider applications and is definitely something that more investors and founders should have in their playbooks.
Why does gaming work?
I would argue that the gaming is an industry that possesses unique characteristics that make it the perfect training ground for technological innovation. This stems from gaming’s unique position as a melange of technology and art, which allows even the smallest technological improvements to create value to the customer, where they otherwise wouldn’t.
A way of illustrating this is by comparing a pure simulation tech company with, for example Improbable.
A pure simulation company that has for instance built out software that looks at the impact of changes in roads taking into account only traffic data, would likely struggle to justify expanding its software to also take into account additional factors (such as the school run, or events) apart from those deemed particularly material (e.g. pedestrians, traffic lights, road works). This difficulty stems from the fact that the incremental improvement to an enterprise buyer (say a government department) would seldom justify the cost of developing such improvements by the software companies. These improvements, would likely be prohibitively expensive in an enterprise or consumer context, and each individual improvement would likely be immaterial.
Improbable’s approach, however, has some significant advantages. Because games are art-tech and compete as such, there is value in incremental improvement insofar as it furthers the art. It is this phenomenon that allows game developers to successfully monetise annual iterations of the same game with only some minor tweaks or slight graphical improvements. When a consumer purchases FIFA 2018 having purchased the 2017 version the year before, they are not purchasing the incremental technological improvement (that would be too miniscule to make sense), they are in fact purchasing a better form of art.
The fact that the art is improved by incremental technological advance is an incidental fact to the consumer but understood by developers and publishers. This fundamental calculus by the business side of the gaming industry creates a perpetual arms race in the industry as companies attempt to improve their art by way of tech. Consequently the industry, able to monetise incremental and even marginal technological innovation as art, will much more happily pay for it than other sectors. Making it the perfect industry for tech companies attempting to develop deep tech capabilities.
Improbable by initially targeting gaming and MMO l developers took the ramp rather than scaling the mountain. This allowed the company to develop its simulation platform able to point to a captive market for each incremental improvements where a traffic simulation company likely couldn’t. While perhaps the traffic simulation company would have been able to aggregate enough marginal improvements to make its product much more compelling, the journey to this ‘critical mass’ of improvements is more expensive and more likely to fail a short term cost benefit analysis. In the best case scenario for our traffic simulation company, it faces a significant time lag between technological improvement and validation where a gaming focused company would experience validation as it rolls out capabilities.
A compelling example of this concept at work in reality is the use of GPUs in artificial intelligence. One could make a compelling argument that the reason Nvidia’s technology was mature and sophisticated enough to be suitable for the vector computing requirements of modern AI was largely due to the fact that gamers’ desire for greater graphical fidelity provided the necessary market allowing Nvidia to continue the improvement of their hardware. I would argue that if Nvidia’s principal market was AI, the company would likely not have reached the heights it has currently achieved, primarily due to a lack of market for an early stage iteration of its product. In this sense AI (and Nvidia) has benefited from the perpetual market for iterative improvement provided by gaming.
I consequently implore, prospective technology entrepreneurs seeking to build out deep capabilities over a relatively protracted period of time to seriously consider using the gaming industry as the incubator for their nascent technology. It is clearly better to develop an advanced physics engine for a market of those obsessed with the reality and fidelity of in game physics for the sake of art than for a property developer or manufacturer obsessed with the strength of his/her bottom line.