AI is advancing rapidly within the enterprise — by Gartner’s count, more than half of organizations already have at least one AI deployment in operation, and they’re planning to substantially accelerate their AI adoption within the next few years. At the same time, the organizations building and deploying these tools have yet to really grapple with the flaws and shortcomings of AI– whether the models deployed are fair, ethical, secure or even explainable.
Before the world is overrun with flawed AI systems, IBM is aiming to rev up the development of open source trusted AI workflows. As part of that effort, the company is joining the Linux Foundation AI (LF AI) as a General Member.Read more over at ZDNet.
I think this is a good step by IBM and it further proves that they are committed to the open source model. This should put to rest the concerns many has over the future of an open Red Hat under the ownership of IBM.
With AI growing more and more in the industry it is only a good sign that the development of open source models and frameworks continues to grow. However, there are still a lot of shortcomings when it comes to open source and AI. The models are only part of a complete AI system. The other, perhaps more vital part, is the data used to train the models. No data, no AI.
For this purpose there are some open AI datasets, such as Skymind.ai, but it is unlikely we will see companies and organisations sharing their datasets as it would give away a lot of their competitive edge. That’s not to say that open sourcing the AI models is a bad thing. But until we get proper data sets open to the public, we will all be in the hands of the companies owning the trained models. Lacking the kind of transparency and freedom that usually comes with the open source model.
Featured image via www.vpnsrus.com