Blockchain Source Code Modification Tracking for AI Black Box
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current AI systems, particularly those using black box technologies like deep learning, face challenges in detecting and tracking source code modifications, making it difficult to understand and interpret changes, which can lead to vulnerabilities such as biased data and incorrect decision-making.
Innovation Solution
A system utilizing blockchain distributed ledger technology (DLT) with plug-ins that record and track source code changes within AI black box applications, providing a transparent history of modifications, including additions, deletions, and reorganizations, and their results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If black box AI systems are used to automatically learn and adapt from data, then the system's ability to solve complex problems improves, but the ability to detect and track source code modifications deteriorates
Solution Approach 1:
The patent introduces a blockchain-based intermediary system that mediates between the black box AI system and the source code. The blockchain ledger records all source code modifications made by the AI system, creating a transparent audit trail without interfering with the AI's automatic learning processes. This intermediary layer allows the opaque AI system to operate autonomously while simultaneously providing visibility into its modifications through the blockchain's immutable record-keeping capability.
2Adaptability or versatility
If deep learning techniques are used to process huge amounts of unstructured data, then the system's problem-solving capability improves, but the transparency of internal operations deteriorates
Solution Approach 1:
The patent implements a feedback mechanism where the blockchain system continuously monitors and records the AI system's internal operations and source code modifications. This feedback loop provides transparency by capturing what the AI system does with the processed data and how it modifies the source code, allowing external observers to audit the system's behavior without limiting its adaptability in processing unstructured data.
3Productivity
If AI algorithms are allowed to make autonomous decisions and modifications, then system efficiency improves, but system reliability and interpretability deteriorate
Solution Approach 1:
The patent applies preliminary action by establishing the blockchain recording system before the AI algorithm executes autonomous decisions. The blockchain is pre-configured to capture and store all modifications and decisions made by the AI system. This preliminary setup ensures that as the AI operates efficiently and autonomously, every action is automatically recorded and made interpretable, thereby maintaining reliability without compromising productivity.
Data Source
AI summary
A system and method for detecting and tracking source code changes in an artificial intelligence black box application using a blockchain DLT. Each block of the blockchain DLT incorporates a plug-in that is triggered when a source code change is made within an AI black box during program execution across an AI platform. Each time the plug-in is triggered indicating source changes are being made by the AI platform's black box a detection and tracking of such changes is recorded in the blockchain DLT as a new block. The specific source code changes are copied and written directly into the new block created and inserted into the blockchain DLT. At the same time, the number of source code modified (i.e., added, deleted and/or reorganized) are determined (i.e., counted) and relevant information identified specific to the modified source code is recorded.