Leveraging Open-Source AI

Open-source software development has been around for a while now, with several popular successes such as the Apache server software, Firefox web browser and LINUX operating system.

More importantly, it has promoted a mentality of consistent improvement of software tools through sharing ideas, refining them collectively and distributing the results. As the fields of artificial intelligence (AI) and machine learning (ML) grow, they could become more intertwined with open-source development.

We have already witnessed Google providing open-source machine learning platforms like TensorFlow with a wholesome ecosystem of tools, libraries, and educational resources from their ML experts.

There are also lots of other free and open-source AI and ML tools such as IBM Watson, OpenNN, Scikit-learn, Apache Mahout, Torch and H2O.

So, with a lot of the necessary elements for this intersection already available, let us examine the impact that the open-source world can have on AI and ML technology:

Open-Source AI in Business

Staff gets to interact with contributions from people in other places, companies, and geographic locations, so their creativity is less constrained even with projects where the goal is quite narrow. They can be inspired and influenced by ideas from other developers.

On the other hand, development in AI and ML that is supported by a platform or set of tools limited to a particular company may restrict growth.

Not only is this approach detrimental to the creation of feedback loops when solving problems, it also does little to develop talent and instill trust in any resultant models and projections.

With an open-source approach, such goals can be accomplished quicker, with less worries about challenges like a lack of talent or high licensing fees.

How do software vendors benefit from sharing their AI technology?

For starters, these companies get to have improved versions of their software at a much lower cost. They also get to be some of the first entities to discover new ventures built on the advancements in their AI technology and possibly invest in them at an early stage.

Organizations like Goldman Sachs have shared some of their pricing and risk assessment code hoping to get new uses for it and use them to earn the support of computer-driven “quant” traders.

When a company shares its AI technology, some of its customers may benefit from the new improvements from the open-source community and continue with the culture of openness.

This may make them more likely to support the company’s entire movement and recommend their products to other businesses.

 

How do companies remain unique amidst all the new sharing?

With open-source AI, companies must no longer spend resources on off-the-shelf AI software that they can barely tweak. For those that already have all the necessary data to drive these tools, they can freely build models at a pace that enables them to keep up with changes in the data.

This makes them become more of contributors to the software and not just users, with their efforts driven by their own business needs. Whenever an improvement is made available, all those to whom it is relevant can obtain it and start at the same level then advance based on their specific field.

The move to open-source AI can also lead to a decentralization of talent, meaning that as more spread over different regions interact with the technology, its adoption rises. When the resultant products become more of a standard in certain disciplines and organizations, this can be good for vendors.

Are you looking to harness the power of open-source communities to advance your organization’s use of AI and ML?

Let the experts at ASB Resources guide you on the different ways to participate in open-source development relative to your organization’s needs. Schedule a call with one of our experts today!