What Can AI Do for Engineers?
AI offers numerous benefits to engineers, but there’s a catch—it can hallucinate. How can you ensure that the answers provided by AI are accurate and reliable? Do you want to cross-check the references AI uses?
One effective way to leverage AI is by providing it with your own trusted data, ensuring that responses are based on verified sources. Let’s explore how I use Google Vertex AI to achieve this.
A Practical Example
Visit www.guofuchen.com and click on "Ask AI PC." The "Knowledge App" is designed to be highly knowledgeable about engineering standards. Let’s test it with a question:
👉 "Based on API, what is the minimum tube diameter for air coolers?"
The AI responds instantly:
"The minimum outside diameter of cylindrical tubes for air coolers should be at least 25.4 mm (1 in)."
Not only does it provide an answer, but it also includes a reference. Clicking the reference opens up the API code, where Section 7.1.11.1 confirms that the outside diameter of cylindrical tubes should be at least 1 inch.
Another Example
Let’s try another one:
👉 "Please list carbon capture technologies."
The AI lists several methods, such as absorption, adsorption, and membrane technologies. Clicking the provided reference takes us directly to a paper written by GLE, supporting the answer.
AI with Verification—A Game Changer for Engineers
With this approach, engineers can feel more confident in AI-generated results. When AI hallucinates, we can catch it!
More Topics
Thanks for watching this video! More interesting topics can be found at showcase.guofuchen.com.
