Artificial Reasoning

tanmay
Inside Machine learning
2 min readJun 19, 2018

You are sitting in an exit row. You casually look at the emergency guide, it is a combination of images and text. Your brain naturally combines them and presents you with a complete picture of the intended message — open the door in the unlikely event of an emergency.

Source: http://legaltechdesign.com/GoodNoticeProject/2014/04/29/united-airlines-safety-card/

As humans this ability to correlate comes instinctively to us but for a minute, think about how a computer sees the same document. An OCR (optical character recognition) system reads the text. An image recognition model scans the image. Then there is a third system that correlates the image and text to understand the complete picture.

The two approaches for completing the same task seem drastically different. However, if you think about it, the fundamental principles are same, the human process of analyzing the world around us and the approach a machine takes to process complex information, both are based on breaking down the data to its core elements. We are just instinctively better at correlating the information.

In my previous post on Artificial Creativity, I argued how AI systems are challenging our notions on creativity. Even though the current pragmatic AI systems cannot replace our true creativity they can certainly stimulate creative endeavors in humans. It can suggest that bacon goes with cocktail. It is still up to us to decide whether we like it or not (I say “meh” on the whole bacon in cocktail suggestion.)

In this relentless pursuit of artificial intelligence augmented capabilities, one key capability has been somewhat elusive — reasoning. The sheer complexity of correlations needed to present a convincing argument is hard for most humans — let alone AI systems.

That’s just changed. IBM recently demonstrated Project Debater that argued with expert debaters on a multiple topics and won (arguably).

IBM Project Debater

We are now at a point where ability to reason can be taught to a machine. Will it soon make better business sense as we delegate basic reasoning and decision making to these systems? What does it mean for us? Do we become obsolete? Does an old Ford become irrelevant because Tesla exists?

I actually don’t think so.

It is an opportunity for us to evolve to an even higher function where the definitions of our fundamental institutions no longer hold. It frees our cognitive capacity to seek out new frontiers in science, philosophy, healthcare, education— it is a win-win!

===

Sources

Artificial Intelligence: Fact, Fiction. How Enterprises Can Crush It — https://go.forrester.com/blogs/16-11-02-artificial_intelligence_fact_fiction_how_enterprises_can_crush_it/

--

--

tanmay
Inside Machine learning

perpetual student. technophile. gamer. comic book nerd. wantrepreneur. @SCSatCMU. @teppercmu. @ibm. @ibmanalytics.