Building and Scaling dApps with AI and Smart Contracts BBLC Practice Questions (Page 4)

Questions 16–20

1. 16expert · hard

   A decentralized insurance dApp uses AI to process claims by analyzing photos of damage. The AI model is trained on a centralized dataset, and the team wants to improve the model's accuracy while ensuring data privacy and avoiding bias. What is the most appropriate approach?

   • ACollect all user data into a central repository and retrain the model with the combined dataset.
   • BHave the smart contract store all training data on-chain so that the model can be audited.
   • CUse federated learning where multiple parties train the AI model on their local data and only share model updates, then aggregate the updates off-chain and use the improved model for claim processing.
   • DUse a pre-trained model without any further training to avoid bias.
   Reveal Solution Discussion Select an answer first
2. 17foundation · easy

   Which of the following best describes how AI can improve data insights within a dApp?

   • ABy replacing the blockchain with an AI database
   • BBy running complex data analysis directly inside the smart contract
   • CBy processing off-chain data and feeding aggregated insights to the smart contract via an oracle
   • DBy using AI to encrypt all data on the blockchain
   Reveal Solution Discussion Select an answer first
4. 18application · medium

   A decentralized content platform wants to use AI to automatically tag and categorize user-uploaded content. The tags are used by a smart contract to distribute revenue to content creators. The team is concerned about the AI making mistakes that lead to incorrect revenue distribution. What is the best way to handle this?

   • AHave the smart contract call the AI service directly and distribute revenue immediately based on the tags.
   • BUse an off-chain AI service to generate tags, and have the smart contract store the tags along with a confidence score; if the confidence is low, delay revenue distribution until a human reviewer confirms.
   • CUse a fixed set of tags that are manually assigned by content creators.
   • DDistribute revenue equally to all creators regardless of tags.
   Reveal Solution Discussion Select an answer first
5. 19application · medium

   A real-estate dApp wants to provide AI-powered property valuations to users. The valuation model requires access to historical sales data, which is stored off-chain. The valuation result must be verifiable on-chain for dispute resolution. What is the best design?

   • AHave users run the valuation model locally and submit the result to the smart contract.
   • BRun the valuation model off-chain, then submit the result along with a hash of the input data to a smart contract that stores it for future reference.
   • CStore the entire historical sales dataset on-chain and run the valuation model as a smart contract.
   • DUse a centralized server to compute valuations and only display the result on the dApp interface without any on-chain record.
   Reveal Solution Discussion Select an answer first
6. 20foundation · easy

   Which AI technique can help improve transaction throughput in a dApp?

   • AUsing AI to compress transaction data so more fit in a block
   • BUsing machine learning to optimize transaction batching and gas price selection
   • CUsing AI to rewrite the blockchain's consensus algorithm
   • DUsing AI to create new blocks faster than the network allows
   Reveal Solution Discussion Select an answer first