ChatGPT has taken the world by storm: It has set records for the fastest-growing user base and, for some, has signaled the dawn of a new era in artificial intelligence. Some have raised concerns about its future implications, while others have heralded it as a game-changer.

Wherever it leads, there appears to be a new wave ahead, with significant implications regarding machine learning and automation.

“What we’re seeing now is something categorically different,” said Jon Turow (pictured), partner at Madrona Venture Group and former head of Amazon Web Services Inc.’s computer vision product. “That’s really exciting, and feels like a durable change. We have these really large models that are useful in a general way.”

Those models can be applied to many different tasks beyond the specific tasks envisioned by designers. That makes them more flexible and more helpful in building applications than what has come before.

“In a nutshell, that’s why I think people are really excited,” Turow added.

Turow spoke with theCUBE industry analyst John Furrier at CloudNativeSecurityCon, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed where AI is heading, how teams can differentiate themselves, and where Madrona Venture Group sees opportunities for entrepreneurs.

The next generation

Though it’s not yet clear which companies will utilize technology like ChatGPT in the most effective way, some trends are emerging. How should companies best differentiate themselves?

“Teams with galactic capabilities to take an open-source model and then change the architecture and retrain and go down to the silicon, they can do things that might not have been possible for other teams to do,” Turow said, noting a company that Madrona has invested in, which provides AI-accelerated video-editing capabilities.

But other teams can also differentiate themselves in three ways, according to Turow:

• Through prompt engineering, on behalf of their users, shaping exactly how the prompts get fed to the underlying model.
• Through information retrieval, concerning how companies get information about the world outside into models so they can reason about them.
• Through attribution, stating where facts are coming from, much as in a news report or academic paper.

Turow has had a front-row seat building technology, leading the product teams for computer vision AI and Amazon Textract and writing the business product plan for AWS IoT and Greengrass. Given that, how does he see the next five years going in this space and where might there be opportunities for entrepreneurs?

“Five years is a really long time given some of this science is only six months out,” Turow said, but he added that the basics are the same as they’ve always been. “We want what I like to call, ‘Customers with their hair on fire.’ They have problems so urgent they’ll buy half a product.”

Companies want those customers, Turow said, because they will meet you more than halfway.

“We want founders who have empathy for those customers, understand what is going to be required to serve them really well, and have what I like to call founder-market fit to be able to build those products that those customers are going to need,” he said.

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of CloudNativeSecurityCon:

Photo: SiliconANGLE