Welcome to the golden age of

Can Chainlink help solve the AI hallucination problem? Laurence Moroney, Chainlink Advisor and former AI Lead at Google, on how Chainlink implemented a novel technique to overcome the risks of hallucination: “They used several LLMs to have them artificially understand the contents of a corporate action and output it in machine-readable JSON format. Instead of trusting a single prompt to a single LLM, the idea was to have a swarm of LLM-prompt combinations to produce various results. The consensus could then be measured. If they all produced the same result, we could begin to trust it.” This technique was used by Chainlink, Euroclear, Swift, UBS, Franklin Templeton, Wellington Management, Vontobel, Sygnum Bank, and more in a recent corporate actions initiative. The models used LLMs by OpenAI (GPT-4o), Google (Gemini 1.5), and Anthropic (Claude 3.5), but the design is agnostic to any LLM model. The results demonstrated success in an LLM consensus framework for financial data—a significant architecture milestone in the journey to improve the management and dissemination of corporate actions data. As Swift’s Kelli West commented at Sibos 2024, this novel solution helps solve “one of the most complex and costly unstructured data problems in the financial world.” Read the full report:

There are currently 11 MILLION crypto coins 99% of them are useless and have no fundamentals The real challenge isn’t finding a crypto project, it’s finding one that actually survives the next cycle