The Pyramid of Artificial Intelligence

1359

We are on the ‘ground floor’ of the Artificial Intelligence Pyramid (and no one knows how to get to the next floor)

I have the great opportunity of hearing between 5 to 10 pitches given by startups every Friday at Plug and Play Tech Center. Roughly for a year since October 2014 when I joined the EIR program at Plug and Play, one in every three startup would mention that they are relying on Big Data. But during the past 6 months artificial intelligence (AI) has taken over the place of Big Data in the pitches.  The problem is that I think in contrast to Big Data the hype around AI is getting way ahead of the real progress.

Despite the hype I’m optimistic about the future of AI and especially about the opportunities for startups to take advantage of the recent progress in these technologies to solve meaningful problems and make positive impact.

To do my part, I’m co-organizing two meetups in the San Francisco Bay Area called Million AI Startups and Artificial Intelligence in Enterprise. During the recent kick-off meeting for the Million AI Startups a number of people requested that before we get into discussing what problems could be solved with AI, we all need to get on the same page about what the current capabilities of these technologies are and what else is coming. This article is my attempt in providing a common language for understanding what we mean by AI so that we can open up the conversation about its potential applications to a wider audience.

AI means many different things to different people, but for the purpose of making sense of how the recent progress can lead to novel solutions I focus on three aspects: Learning, Autonomy, and Reasoning. These aspects (roughly) correspond to the three main school of thoughts about artificial intelligence, namely: Connectionism, Dynamicism, and Symbolicism.

Luckily with three aspects or dimensions  we can visualize a 3-dimensional space where existing and future AI technologies and solutions may be placed and compared with one another. To stay pragmatic, the scope of AI technologies that I am discussing here is limited to human-like intelligence and therefore the 3D space is enclosed by what I call the Pyramid of Artificial Intelligence that is depicted below.

read source blog on Linked In by Cirrus Shakeri