Big Data: Cost Control May Be a Challenge
December 24, 2018
I read “AI’s Dark Secret? A Desire for Data.” The write up states:
The AI revolution is hungry for personal data.
Those data come with a catch.
To ensure that AI algorithms work properly and to get the bugs out, they need to fed a consistent stream of data. The data needs to be reliable, accurate, and objective and that costs a lot of money. Venture Beat shares how data has a downside in the article, “Could Data Costs Kill Your AI Startup?”
AI startups that discover their funds are chipped away by data costs should consider moving that cost from the research and development line to the costs of goods sold column. The article explains it is a golden opportunity to scale up your company, drive costs down, so that margins will increase.
Startups use data in three basic ways: acquiring, storing, and annotating the data to train the algorithm model. All these steps cost money and can tack on more expenses based on what resources and services you offer. There are different ways to scale down costs at each of the steps, but how and what depends on your individual project. The best way is to figure out how to optimize not only your costs, but also all of your tools:
“The first successful AI businesses came to market offering AI-free workflow tools to capture data that eventually trained AI models and enhanced the tools’ value. These startups were able to achieve software margins early on, since the data and AI were secondary to the startup’s value proposition. As we move to more specialized applications of AI, however, the next wave of AI startups will face higher startup costs and will require more human labor to provide initial value to their customers, making them resemble lower-margin services businesses.”
The only fact you can be sure of with your AI startup is that costs will continue to rise. In order to maintain your relevancy and sell your product, figure out how you can make the most of everything available to you.
Whitney Grace, December 24, 2018
Comments
-
- Subscribe to Beyond Search
Feature archive
News archive
-
