There’s no denying it — Artificial Intelligence is happening and it’s happening big. Companies from Facebook to Google to Amazon are hard at work building world-class AI teams that infiltrate every facet of their products. Siri is one of the largest teams at Apple, and Microsoft has a growing research effort on this front. But why is now the time for AI?
Here are 9 reasons AI is happening now:
1. Artificial Neural Networks
Traditional programming is deterministic, sequential, and logical. For example, computers take inputs, apply instructions, and generate outputs. This is great for tasks like calculations and conversions, but ill-suited if the application isn’t explicitly defined. The human brain, on the other hand, doesn’t behave this way. We learn and grow through repetition and education. Recent progress in artificial neural networks (ANNs) is key to building computers that can think. These breakthroughs are enabling tremendous strides in AI work at Google and Apple.
In machine learning and cognitive science, artificial neural networks (ANNs) are a family of statistical learning models inspired by biological neural networks (the central nervous systems of animals, in particular the brain) and are used to estimate or approximate functions that can depend on a large number of inputs and are generally unknown. Artificial neural networks are generally presented as systems of interconnected “neurons” which send messages to each other. The connections have numeric weights that can be tuned based on experience, making neural nets adaptive to inputs and capable of learning. Wikipedia
2. Knowledge Graph
Companies like Yelp, Foursquare, and Wolfram Alpha have enabled access to their data through APIs. As a result, platforms like Siri and Google Now are able to answer questions such as “What’s the closest coffeeshop?” or “What’s the population of India?”. If a new service had to handle the natural language processing (NLP), audio processing, data, and more, it would be nearly impossible. Fortunately the knowledge graph has evolved over the last 20 years to a point where new AI platforms can immediately have access to tons of data.
3. Natural Language Processing
NLP is a field of computer science and linguistics where computers attempt to derive meaning from human or natural input. While the field has been around since the 1950s, we’ve seen huge strides in the last few years thanks to Markov Models and n-gram models as well as projects like CALO and Wordnet. Stanford’s CoreNLP (demo here) is one of the many strong NLP solutions available today:
4. Speech Processing
In order to speak to a computer and have it understand our intent, we first need to handle the audio processing and convert sound waves to text. Known as speech processing, this field has seen major advancements in the last few years. Beyond the advancements in technology, we’ve seen companies like Nuance emerge with powerful APIs that power services like GPS, dictation, and more. Today, it is almost effortless for a new AI company to translate voice to text with a high degree of confidence.
5. Computational Power
The increase in computational efficiency over the last 17 years has been remarkable. In 2014, people could buy a video card that was 84.3 times the performance of one from 2004 for the same price. This increase in computational power is necessary if we want to emulate the brain. For example, research attempting to simulate 1 second of human brain activity required 82,944 processors supporting 1.73 billion artificial neurons connected by 10.4 trillion synapses. The decrease in cost and increase in computational power is enabling tremendous breakthroughs in AI today.
A big aspect of seeing mass adoption around artificial intelligence is consumer approval. With an initial push from Apple to highlight Siri, and now Microsoft’s Cortana and Google Now doing the same, smart phone owners have access to an AI whether they like it or not. As a result, consumers are coming around to the idea and even starting to embrace it. Funny videos like this one are helping the masses to accept this new human-computer interaction:
7. Ubiquity of Personal Computing
Conversing with an AI is a very personal experience. The emergence of smaller, always-on devices makes this possible. The iPhone was first introduced in 2007, only 8 years ago. Now, more than 64% of Americans own a smartphone. Wearables, such as the Apple Watch or Jawbone, open the possibility of even more intimate personal computing. These devices that we carry or wear serve as excellent hosts for this technology, making it possible for AI to truly enter the mainstream for the first time.
AI funding seems to go through waves, and in the last few years it’s definitely back up. Scaled Inference, a predictive AI company, recently raised $13.6M. Amazon just announced a $100M fund for voice controlled technologies, and IBM did the same for the Watson Venture Fund. The total invested in AI companies in 2014 grew past $300M from a mere $14.9M in 2010 according to Bloomberg. With firms like Khosla Ventures and Andreesen Horowitz leading deals in AI companies, funding is fueling innovation in AI.
9. Research Efforts
Another reason for the apparent surge in AI is the collective research efforts taking place. According to a 2014 report by MIRI (Machine Intelligence Research Institute), 41 of the top 275 CS conferences are AI-related. AI accounts for about 10% of all CS research today. The IEEE Computational Intelligence Society has more than 7,000 members and there are more 106 AI journals. Based on MIRI estimates, more than $50M went into funding AI research by the National Science Foundation (NSF) in 2011. With this much research and effort going into AI innovation, it’s no wonder we’re seeing this technology starting to reach the masses.
If history is an indicator, we may see interest in AI spike and go back down. With momentum across these various different sectors, though, AI interest seems likely to keep growing. If you’re interested in keeping up with our efforts and staying in touch, check out http://josh.ai and reach out!
This post was written by Alex at Josh.ai. Previously, Alex was a research scientist for NASA, Sandia National Lab, and the Naval Resarch Lab. Before that, Alex worked at Fisker Automotive and founded At The Pool and Yeti. Alex has an engineering degree from UCLA, lives in Los Angeles, and likes to tweet about Artificial Intelligence and Design.