Austin’s Lucid Aims to Make Artificial Intelligence More Human

Austin — Many of the advances in artificial intelligence have focused on the numbers, algorithms, and patterns that help computers to predict human behavior. An Austin, TX, company wants to connect that approach to one that uses more human-like reasoning, as well.

“The ‘thinking like a human’ part is much more difficult to build,” says Michael Stewart, CEO and co-founder of Lucid. “It takes 20-something years to organically build a common-sensical human. How do you do that in computing?”

Stewart says Lucid has figured out how to combine the mathematical analysis in most artificial intelligence software with the ability to answer the questions “why?” and “so what?”

Lucid’s target customers are firms in industries like finance and healthcare that, Stewart says, especially to make decisions based on a deeper understanding—software that not only tells them what, but also why. Right now, he says these businesses have largely mined what they can from current “left-brained” data analytics. “They’ve exhausted the use of statistical tools,” he says. “They all have them and they find the same things.”

For example, consider a hospital. “In genomic research, you can use math to understand the probabilities of a disease off of a certain genomic activity,” Stewart says. “The problem is once you find the potential for the correlation, you have to have a bunch of humans evaluating the causation and the relevance of a drug to solve that.”

For one of its customers, the Cleveland Clinic, Lucid says its software enabled doctors there to better sort through 5 million patient records to find clinical trial candidates—10,000 times faster than with traditional approaches.

While Lucid is a young company—it was founded in 2008—its origin story veers from the traditional startup path. The company’s technology—called Cyc, for “encyclopedia”—was originally developed by Doug Lenat, a former Stanford University professor lured to the Microelectronics & Computer Technology Corp. or MCC, a research consortium founded by 12 technology companies in 1983.

In an interview with the Austin Chronicle in 1999, Lenat, who is now Lucid’s co-founder and chief scientist, described Cyc as a repository of “obvious” knowledge. “Knowledge that is so obvious that it’s confusing or insulting to ever say it to someone else,” he said. “Knowledge like if you’re holding a glass of water, you should hold it with the open end up.”

Stewart came on board in 2008 and has realigned Lucid’s focus to be on the healthcare and finance industries, while also serving its older government customers, such as the U.S. Forestry Service.

Investments into the company have come from people such as former Dell CFO Tom Meredith. Stewart declined to provide more details about the company’s fundraising.

In addition to expanding the business, Lucid has also formed an Ethics Advisory Panel to provide ethical guidelines as the company deploys its artificial intelligence software for customer needs. The panel includes artificial intelligence experts such as Murray Shanahan, a professor of cognitive robotics at Imperial College in London, and MIT physics professor Max Tegmark.

“AI is going to be a profound disrupter to humanity, to all activities; it has the potential to be negative like any powerful tool,” Stewart says. “If you want to be ethical 10 years from now, you probably should be ethically focused and obsessed early.”

Angela Shah is the editor of Xconomy Texas. She can be reached at ashah@xconomy.com or (214) 793-5763. Follow @angelashah

Trending on Xconomy