IBM and Casper Labs will collaborate to help companies better understand artificial intelligence systems.
The partnership of the mentioned brands was announced last Thursday, January 11. Casper and IBM Consulting will develop a solution that will create an additional analytics and policy enforcement layer for managing data that is used in the training of artificial intelligence systems.
The press release of the parties to the cooperation notes that the process of training, developing, and deploying generative models of machine intelligence takes place in several organizations. In this case, the creator of the AI configuration acts separately. Then, an end-user organization works with the artificial intelligence model.
The press release also notes that as various firms integrate new datasets or modify AI configurations, the output information changes. Against this background, according to the parties to the cooperation, many organizations need to be able to track and audit the specified changes. Companies also need tools to accurately diagnose and remediate issues.
The new solution will be based on the Casper ledger, which is protected from unauthorized access and is highly serialized. IBM watsonx.governance and IBM will be used to monitor and measure highly serialized input and output data for training artificial intelligence systems in organizations. watsonx.ai. The relevant information is contained in the press release of the companies.
The hybrid nature of the Casper blockchain will help organizations improve the protection of confidential data from access by external actors. Also, in this case, consumers will be able to control who can get acquainted with the arrays of information.
Shyam Nagarajan, global partner, blockchain, and responsible head of artificial intelligence at IBM Consulting, says that the effectiveness of an artificial intelligence system is proportional to the ability of an organization to manage it. According to him, firms need solutions that strengthen trust, increase explainability, and mitigate risks.
As the process of using artificial intelligence systems in corporate spaces scales, IT services, data processing services, and security services must improve their understanding of the risks inherent in each stage of AI development.
The open exchange of information arrays is a major component of training digital intelligence configurations. Researchers collect and share data to create a kind of educational material for artificial intelligence models. Sharing large amounts of information increases the risks if the relevant process is organized incorrectly. In this case, an important circumstance is that artificial intelligence can disrupt almost all the internal mechanisms of the functioning of organizations within a single action.
In the context of using AI in a business environment, the main problem is the integrity of the data that is transmitted to the advanced technology model. To use artificial intelligence configurations efficiently and safely, the right information infrastructure should be created.
As we have reported earlier, IBM Expands Collaboration With AWS.