IBM has announced the launch of AI OpenScale, a new service which will aim to provide companies with transparency and remove bias from AI models.
AI OpenScale will address the lack of visibility within AI models, while detecting unwanted bias and interoperability among tools and frameworks by providing explanations into how AI systems make decisions. It also detects and mitigates bias to produce fair and trusted outcomes.
The AI OpenScale platform will attempt to bring confidence to companies that are just starting out on AI projects by addressing the challenges that come about when first adopting the technology. It is intended to act as an intermediary layer between the application and ML models delivering predictions.
The service is now available within the IBM Cloud Watson.
"Innovative organizations understand that AI offers competitive advantage and is already a driving force in the evolution of their industries," stated IBM in a blog post announcing the launch of AI OpenScale.
"However, for every successful implementation, there are many AI models that fail to make it out of the lab and into production. At IBM, we understand the roadblocks to full AI adoption – from trust and transparency concerns with the "black box" of AI, to problems of scale and automation – and we are working with enterprises to provide the tools needed to overcome these roadblocks and deliver real business value," IBM added.
AI OpenScale is the latest in a number of AI projects IBM has launched over the last few months. In September, it launched an AI toolkit to avoid bias and an AI app to drive farm and factory business growth, in addition to several other strategic moves utilizing AI and other disruptive tech. It was also recently revealed that IBM had applied for the second most blockchain patents as a company, just behind Alibaba.
But despite IBM's recent moves, last week it was reported that it has lost 12% of its value this year alone. IBM chief financial officer Jim Kavanaugh stated, however, that the business's financial standing was expected to improve and that the losses this year were due to IBM "wrapping up (its) most successful mainframe product ever".