Orions Systems | 2016: The Year of Artificial Intelligence
17542
post-template-default,single,single-post,postid-17542,single-format-standard,ajax_fade,page_not_loaded,,qode-child-theme-ver-1.0.0,qode-theme-ver-10.0,wpb-js-composer js-comp-ver-4.12,vc_responsive
 

2016: The Year of Artificial Intelligence

2016: The Year of Artificial Intelligence

Nils Lahr, CEO and Founder, Orions Systems

If there’s one technology story that defines 2016 I believe it was the acceptance of artificial intelligence as a part of our everyday lives. As a computer vision company and one of the earliest entrants in the field of AI, we’re aware that for some the term “artificial intelligence” carries negative connotations, on the spectrum from skepticism to fear.

But this shifted in 2016. Skepticism is giving way to cautious optimism and fear is decreasing through more frequent contact with AI in our everyday lives. Amazon’s Alexa, trojan-horsed into our homes via the company’s Echo device, is paving the way for a variety of computer-based devices and home robots. The advancements in natural language query combined with instant access to endless data and algorithms power snappy, intelligent responses to our requests.

In 2016 autonomous vehicles left the test parking lots and entered the roadways, soon to be a mainstream technology.

AI even made moves into the creative space. IBM’s Watson analyzed all the scenes from a movie, pulling the ones most likely to be used in a trailer, and with the help of human editors cut the time of trailer creation from several weeks to several days. An algorithm that told Watson different elements from scenes that have been used in trailers in the past drove the scene selection.

Compute Power from the Cloud

The shift taking place is driven by new levels of compute efficiency making it possible to process data at a scale that business rules can be applied and optimized. Particularly with heavy data such as video and imaging, the scaling power of the cloud has been a breakthrough element in driving the machine-learning loop that’s necessary to create more accurate intelligence, training and optimizing the algorithms behind these experiences.

Not surprisingly each of the major cloud providers, Microsoft, Google, and Amazon made big AI investments in 2016, together and going it alone.

As data continues to grow, and in our case video data which is growing at a tremendous rate, it only becomes more difficult to find the signal in the noise that feeds real intelligence back into the cycle to create smarter business rules and algorithms. The sheer compute power needed to just ingest, store, manage, and process this amount of data has made it cost prohibitive to most industries or lines of business. But decreasing compute costs driven by increasing cloud-usage are driving down costs opening up new markets and opportunities to use AI and computer vision solutions to find intelligence in video and image data.

Asynchronous AI and Human Workflows

The compute power of the cloud is one element of technical infrastructure that’s supporting AI. Another element is the ability for humans and machines to work together.

Even as we’ve seen vast improvement in AI we’re reassured that people are still very much in the driver’s seat. It’s human expertise that selects and feeds data to Watson, humans that create the rules that determine the moves of an autonomous car, and in our case it’s humans working in parallel with computers that provide the level of intelligence and accuracy that computers simply can’t on their own.

In the early 2000s, Palantir, the highly valued big data analysis company, took a hybrid approach to developing their first solutions for the CIA using technology to empower experts from within intelligence agencies to more intelligently explore their data, a goal of intelligence amplification or augmentation, or IA vs AI.

“Palantir see human intervention, whether in the form of a Data Scientist or other specialists as critical. Palantir’s view is refreshing as it highlights the complexity of advanced analytics and the need for human intervention to ensure insights are rigorously validated.” Ntegra Greenside IT Director Study, 2014

 

In 2016 we are seeing examples in computer vision and in other AI solutions where humans and machines are working together to drive the machine learning loop faster.  As impressive as the investment funds from the major tech players are, it’s industry standards, open systems, and sharing of data between organizations, human experts and machines, that will determine the speed at which AI develops over the coming year.  We’re excited to be a part of this next wave of growth in AI.

 

nils-headshotNils Lahr is the founder and CEO of Orions Systems, providers of Orions Smart Vision Platform, a hybrid solution for unlocking intelligence in high value video.