.png)
The democratization of AI and it's rise in Construction Startups
Author:
Priyank Savla
May 13, 2019
In 2019, saying that you are leveraging AI is like saying you are leveraging the "cloud." An AI company is like saying you are a "Cloud" company. AI is simply a tool. Your primary focus should be how well a technology solves a problem, regardless of the use or degree of use of AI technology. It seems that these days, each new construction startup is using "AI" in some way. Where is all the emphasis coming from? One of the biggest reasons is that various parts of AI pipelines are quickly becoming commoditized as a service.Remember when "cloud computing" came out? It meant that you no longer had to hire full-time database engineers to maintain your servers since you could now easily "rent" cloud services from Google, Amazon, or Microsoft. Companies who had thousands of engineers building servers at a scale. Shortly after that, we saw the rise of many large companies such as Salesforce and Procore who could focus their efforts on solving customer problems, while on-premise solutions which failed to transition to the cloud became obsolete. The same trend is happening to AI now, which is facilitating growth and use of AI technology across all businesses. This adoption allows construction startups (such as ourselves) to focus more on solving the problem at hand, delivering the best user experiences as opposed to trying to figure out how to deal with the plethora of nuances that accompany building an AI pipeline and it's underlying technologies from scratch.
There has been a commoditization of Cloud technology as a service by big companies like Amazon, Google, and Microsoft. It's no longer competitive to hire an engineer to maintain your internal networks. The same thing is happening with AI. CEMEX Ventures has backed us for this reason: they believe that it's easier to teach people who genuinely understand the business problem how to leverage AI, versus teaching technology people how to do construction.Examples of enabling technologies include:
Unless you are solving a problem that requires leveraging AI in a very unique way that is not efficiently supported by these services, it may be seen as a competitive disadvantage to have too many AI-focused engineers on your team who may be "re-creating the wheel." Instead, leveraging a service that is being continually refined and developed at-scale by thousands of AI engineers/research scientists at Google, Amazon, Microsoft, or IBM.
The benefit in outsourcing is the money you could have spent on a top-of-the-line AI/ML scientist can now be allocated to hiring a fantastic web/back-engineer to efficiently build a pipeline using a series of micro-services and more actual features that construction customers can benefit. An AI pipeline needs to accomplish 2 things: achieve a business objective, or accomplish a technical objective, and these two things are very different. As an example...[], also, there are many platforms available that create various abstract parts of an AI pipeline; this includes facilitating data cleanup and preparation, to the automatic creation of data models by services like Google Cloud AutoML.
