✕

Sheet metal shops have used a basic form of artificial intelligence for decades now, with nesting software that automatically iterates between different cut patterns until it finds the least wasteful configuration.

This rules and math-based process shows the first form of a technology that is now common parlance in company earnings calls for the potential of more advanced, generative AI.

“There's a balance of companies who are doing more optimization, mathematical reasoning to companies who are more on the offense with predictions and learning,” said Jordan Loyd, partner at Vation Ventures. He recently presented at the Spiral Duct Manufacturers Association meeting on the potential for artificial intelligence to multiply worker output.

“People who use AI will replace people who don’t,” Loyd added. “If you sleep on this, it’s going to take somebody 3 or 4 months to make a product that took you 3 or 4 years.”

While he is a proponent for AI, Loyd cautions that there are already an exhaustive list of different AI solutions out there now – with new ones coming out every day. Picking the right, most market-ready solutions is vital, as is R&D in developing more niche plugins.

The marketing, legal, consulting and education industries are leaders in adopting generative AI, but according to Menlo Ventures, only 1% of manufacturers use generative AI. Lloyd said marketing is an “easy, low risk” area to deploy generative AI to maximize worker output with generative emails, SEO analysis and copywriting.

“If you have one person on your team with an instance of Open AI, that person can do the [work] of tons and tons of people,” Lloyd said, adding legal work is also easily multiplied with generative AI software.

“When we think about contract management, legal contract, lifecycle management with AI, it crushes,” Loyd said. “Think about the things that are agnostic to any business because those are going to be the solutions that are more mature in the market.”

In that vein, cybersecurity and information technology are increasingly adopting generative AI, in part to keep up with security threats that likewise are associated with AI. As many as 55% of data loss protection events from mid-January to mid-February involved users adding identifiable information into generative AI tools, according to a Menlo Security report published earlier this year.

“It's really getting popular in the coding side of things. When we think about developers, IT, cybersecurity, there’s a huge amounts of opportunities,” Lloyd said, noting that long-term, one of the biggest cybersecurity threats will be intentional attempts to pollute models with wrong information.

“It’s important to have your guardrails up in these early stages,” Lloyd concluded.