In 1956 John McCarthy, a professor at Dartmouth college, invited a group of colleagues to gather for a summer workshop. His stated goal was “an attempt will be made to find out how to make machines use language, form abstractions and concepts, solve kinds of problems now reserved for humans and improve themselves. We think that a significant advance can be made in one or more of these problems if a carefully selected group of scientists work on it together for a summer.” While McCarthy’s workshop did not achieve any of the lofty objectives stated above, this is generally acknowledged to be the point in history where the notion and study of artificial intelligence (AI) began. A lot has been accomplished since this workshop, yet there is still a lot to do.
Artificial intelligence is a branch of computer science devoted to developing data processing systems that perform functions normally associated with human intelligence, such as reasoning, learning and self-improvement (ISO/IEC 2382:2015 Information Technology – Vocabulary). AI has a rather long and disjointed history – with many stops and starts since that first get together at Dartmouth. Over the last few years AI has made its latest resurgent into the forefront with great promises to reshape business in the next decade. The reason AI is back in the limelight include:
- Exponential increase in computational power (Moore’s law)
- Explosive growth of data available to our machines – the amount of data we produce doubles every year, it is estimated that in ten years there will 150 Billion networked sensors – if this is the case the amount of data available could double every 12 hours)
- Engineers and computer scientists have shifted the focus of AI systems from those focused on broad applications to solutions that attempt to address deeply a specific area.
- Engineers and computer scientists have transitioned their mindsets away from encoding a large set of rules, but rather towards building systems that learn from exposure to data sets.
AI solutions generally fall into one of two categories
- Narrow (sometimes called weak) AI – this is where most of today’s commercial AI systems reside. These systems are focused on executing a simple task or several tasks of the same genre. Examples applications of narrow AI include question answering and smart home devices. Alexi, Google Assistance are built by combining several Narrow AI solutions together.
- General (sometimes called Strong) AI – this is where we are currently headed. Strong AI systems have the ability to understand context and make judgements based on that context. They can learn from experience, make decisions, use reason and act creatively.
We encounter apps and devices that are employing various aspects of AI solutions every day. Pandora studies the musical tastes of all its users, determines which users are ‘like’ other users, and hooks users up with songs that they didn’t even know they liked. I have found it to be uncannily good at this (except for playing Johnny Cash for me occasionally). Netflix knows what you like to watch, and Amazon know what you like to buy. Automobiles continue to get smarter thanks to AI algorithms – telling you when to break or just simply breaking for you.
These applications of AI are just the start. The closer we get to commercial applications of Strong AI, the more we will find ourselves sharing the work load with computers.
I will be presenting a free webinar on this topic on Thursday April 4, 2019 at 11AM EST. Click here if you are interested in learning more about the history and future of AL