Editor’s note: Artificial intelligence is being called this age’s most profound opportunity or gravest threat. Thought leaders from across North Carolina and beyond gathered last week in Durham to discuss the implications of AI on the future of work and beyond. WRAL TechWire’s co-founder Allan Maurer attended the event, and in this the second of several articles he provides a detailed look at AI from the viewpoint of Triangle companies from startups to global leader IBM. This story focuses on Robbie Allen, whose new company is focused on machine learning.
We are nowhere near a path to the kind of general artificial intelligence the human-like androids of Star Trek, Westworld, and other science fiction would need, said Robbie Allen, founder of two Triangle-based AI-focused companies. Allen, founder of Automated Insights and Infinia ML, said, “We are at the height of the hype cycle when it comes to expectations for what AI can do,” in a TED-style talk at the annual NC Tech event, The State of Tech, last week.
Right now, Allen said, “We can simulate about as many neurons as an ant has, but still can’t do nearly what an ant can do. ” So full automation of all labor is not on the current event horizon, he said. “Humans have hard to automate features.”
Allen’s Automated Insights, which sold to private equity firm Vista Equity and its portfolio company, STATS in 2015, created Wordsmith, the robot news writing software that takes content based on numerical data such as financial earnings reports and sports stories and turns them into human-like narratives.
AI won’t be writing Harry Potter sequels
Associated Press produces thousands of earnings reports each quarter using Wordsmith and Yahoo used it for fantasy sports stories. “You probably read stories written by Wordsmith and never knew it,” Allen told the NC Tech audience.
The AI-based program and others like it gave journalists and other writers a scare, but getting machines to understand natural language, a Holy Grail of AI, “Is extremely challenging,” Allen explained. “
You can program all the Harry Potter novels into a computer but that won’t enable it to write a sequel. What you get back is akin to gibberish. Machine learning and AI are not yet up to the task of understanding complex natural language narratives.
“AI won’t generate books.,” Allen said. It probably won’t be doing narrative journalism any time soon, either.
That doesn’t mean there is not a lot of amazing things happening in the AI, machine learning, and natural language processing space, he noted. “AI is a moving target. It’s even hard for me to stay up to date. But we are in the very early stages and it will take time for the technology to mature.”
The technology has made innovative advantages around image recognition and certain mobility characteristics of robots, he said. Computer vision can do things such as automatically recognizing what’s in a suitcase from an X-ray, for instance. “We’re seeing amazing results and we’ll see improvements the next couple of years,” he said.
Allen’s newest startup,the Duke University spinout, Infinia ML, develops machine learning systems for large businesses. It’s focused on “using machine learning to make a business impact.”
No trouble attracting top talent to the Triangle
The company’s website asks, “What if the next great breakthrough is hidden in your data?” And, “What if security threats were easy to see?”
Infinia ML says its algorithms extract meaning from large numbers of complex documents, helping companies automate mundane tasks that drain time, budgets and attention. It can scan images and videos for anomalies humans might miss. And it unlocks hidden insights in a company’s data.
Allen says Infinia ML has already signed several Fortune 500 companies and is headed for multi-million run-rate for 2018. In an email, Allen said he often hears concerns about whether the Triangle can support and attract top technical talent, but InfiniaML is certainly not seeing that.
“For the first four months of 2018, we’ve had 1,261 applicants for our open positions, including 504 for our data scientist position. That’s more than ten applicants per day on average. We have had a healthy mix of both local and out-of-town candidates.”
He added that the company is not using any outside recruiting help and candidates are coming from word-of-mouth, LinkedIn, or job listing site, Indeed.com.
Allen and two co-authors are writing a book on machine learning for publication in 2019, and he will be teaching a class at the Kenan Flagler Business School at University of North Carolina in the fall on machine learning.
More coverage: Can artificial intelligence bring more truth to marketing?
IBM exec says AI won’t eliminate all jobs – but they will be different.
