Trend Monitoring Resource
Exciting news! Our research lab has made several advances this week. As shown in the attached picture, we have developed a fully automated trend monitoring radar that scans the internet on a weekly basis.
The engineering process was a perfect blend of automation script and AI based on “prompt engineering”. However, there seems to be some misunderstanding around what prompt engineering is, and many are unaware of the academic paper that first discussed it. Prompt engineering is an emerging technique that enables us to “command” generative AI to process our objective tasks precisely as we would like. Currently, there are three broader techniques to consider on this topic: 1) extending scripting and vector databases to leverage techniques such as “chain-of-thought” prompting, “one-shot,” “few-shot,” etc.; 2) combining various prompting techniques (particularly the Transformer Attention Mechanism) to enable the generative AI to perform like a Turing machine; and 3) augmenting multiple LLMs AIs to crosscheck and verify each other.
We will delve into this topic in detail on Monday, and on Wednesday, we will deploy the fully automated trend monitoring radar (codename: “PulsarWave”) and scenario planning (codename: “PlanarNexus”) accordingly. In the future, we hope that Wednesday’s story will be fully automated with our supervision. Stay tuned for more updates!
