ChatGPT is barely a year old, but engineers have already embraced it as an invaluable tool to keep tedium at bay. Here's how you can too.
It’s been a year since OpenAI changed the world by releasing its generative AI tool, ChatGPT. Since then, the number of ChatGPT end-users has skyrocketed to the hundreds of millions, and tech giants including Amazon, Meta and Google have scrambled to release competitors. Countless software vendors have now incorporated generative AI into their tools, and engineers and others have been astounded by the impressive value of these new features.
Engineering.com spoke with several engineers to learn how ChatGPT and other generative AI tools are boosting their productivity. If you haven’t started taking advantage of generative AI yet, these are excellent places to start.
Generative AI as an engineering assistant
No engineer should worry that generative AI will replace them, but every engineer should look for ways it can assist them. Using generative AI as an assistant can make you look brighter, more accurate and more productive. Who wouldn’t like that?
For example, you can confirm details of mathematical formulas, competitor announcements or the specifications of specialized materials faster with generative AI than with search.
“I experimented with multiple generative AI software packages, including Bing and Poe Assistant,” said Mel Head, a retired chemical engineer with a computing background who worked at Honeywell. “I’ve produced the best result when I work with AI, treating it as an assistant, not simply as a new tool.”
Summarizing engineering information
Engineers spend too much time researching and reading lengthy academic papers and barely readable user manuals in pursuit of essential nuggets to advance their work. Extracting those nuggets from extensive collections of documents is an excellent generative AI application.
For example, with generative AI, summarizing a library of internal reports is faster and produces more consistent results than assigning an engineer-in-training to this task.
“I use Sharly to rapidly find and summarize engineering information stored in various files, including PDFs, text files, Word documents and other electronic formats,” said Brad Henrie, quality and process engineer at Sealweld Corporation. “Sharly is saving me significant amounts of research and reading time.”
Developing software
Many engineers are involved with software development to some extent. Software development is creative when engineers tackle new problems, but it is tedious because of the minutiae today’s programming languages require. Generative AI is excellent for producing first-draft software at astonishing speed. Engineers must carefully review and test the generated software to ensure it produces accurate results—something AI can’t do.
For example, using generative AI to produce a first-draft software application for managing IIoT data from a plant is faster than coding it by hand. That’s because the engineer no longer has to code all the repetitive input-handling, data validation and read-and-write operations.
“ChatGPT is a powerful tool for engineers to advance work on software development rapidly,” says Damien Hocking, CTO at Madala Software. “AI tools don’t give you a perfect solution, but it’s fantastic as a flying start that saves hours of grunt work.”
Creating written content
Most engineers view writing reports as a painful task that distracts from the more engaging and creative engineering work. Generative AI happily cranks out impressive first drafts, while engineers only have to polish and finalize their report.
For example, generative AI can quickly produce reports from lab notes or text for web pages, emails or engineering case studies.
“With Airtable and ChatGPT, I can jot down 100 short topics,” says Alan Mourgues, consulting reservoir engineer and founder at CrowdField. “When I’m back from my coffee break, I can scroll through 100 draft blogs ready to go.”
Comparing engineering regulations
Regulations can be voluminous and complex to understand. Looking for differences, omissions and contradictions across multiple jurisdictions is an impossible manual task even for a small set of regulations. Generative AI can sort through the texts from many jurisdictions and present engineers with highlights, gaps and incompatibilities.
For example, generative AI can quickly compare regulatory documents for differences in text without being distracted by slightly different terminologies or synonyms across various documents.
“I use ChatGPT and Bing Chat to save me time researching information and comparing regulations across jurisdictions,” said Mark Perrin, petroleum engineer and VP at TriAcc Group. “Bing Chat is more useful to me because it’s more tightly integrated into our Microsoft Windows environment.”
Drilling into information
It’s almost impossible to type an initial generative AI prompt with sufficient accuracy that the first response provides precisely what you need. Generative AI is not limited to a single prompt and response. Continuing the dialogue to refine or clarify responses makes the result more accurate and targeted for engineers.
Generative AI shines when processing text. Engineers can use models such as Anthropic’s Claude 2 to first summarize lengthy reports in minutes. Then they can continue a “conversation” with the generative AI about the contents by drilling down into issues of interest, which can save a lot of time.
For example, you could start with a broad prompt such as: What types of paint can protect steel? Then, you could narrow down to the desired result through successive follow-up prompts such as: What type of paint is optimum for a marine environment?
Don’t forget human judgement
2024 will bring a new wave of AI-based software tailored to specific industries and individual professions, including engineers. This software will help every engineer improve the quality of their work while reducing effort. However, don’t forget to always review generative AI results for accuracy and continue to exercise critical thinking, because AI doesn’t offer that. Sometimes, AI results are inaccurate or misleading imitations of critical thinking.
—
Yogi Schulz has over 40 years of Information Technology experience in various industries. He writes for ITWorldCanada and other trade publications. Yogi works extensively in the petroleum industry to select and implement financial, production revenue accounting, land & contracts, and geotechnical systems. He manages projects that arise from changes in business requirements, from the need to leverage technology opportunities and from mergers. His specialties include IT strategy, web strategy, and systems project management.
