Despite the distance that AI has come over the last years, it’s still in its infancy stage. And despite that fact, it’s creating a massive drift in all fields of civilization, nowadays. As Sam Harris defined it, Intelligence is but a matter of information processing, and the only path we have as a civilization is to continue improving our information processing ability, which eventually will create intelligence that it is greater than humanity combined. This process involves creating both, highly capable Artificial Narrow Intelligence systems and Artificial General Intelligence systems. We’re currently approaching the perfection of Artificial Narrow Intelligence, that is, systems that are extremely intelligent performing a certain specific task. The jump from Narrow to General is huge, but we’ll eventually get there. So let’s explore in more details the ventures of both technologies across various fields.
Artificial Narrow Intelligence:
Artificial narrow intelligence is currently an industry-leading technology in almost every single field, the cutting edge people have sought for ages. One rising question is whether Artificial Narrow Intelligence will drive a massive unemployment wave. The short answer is yes, but it’s going to create meaningful jobs as well. It’s excessively imprudent to say that some jobs are better off done by humans, but for a close-visioned task doer like an artificial narrow intelligence, you might need a human presence to avoid any abnormalities that occur out of the system’s scope.
In a famous study, it was shown that around 47% of Americans are at the risk of losing their jobs to automation and AI. Let’s start by showcasing the advances of AI so far.
Arguably, the most breathtaking accomplishment of AI so far is AlphaGo, as it evolved through 3 versions, Lee, Master, and Zero. The latest managed to reinvent itself without the need for humans. AlphaGo Zero learned the game only by playing itself, and in a matter of 3 days, it reached Lee’s level and 21 days to Master’s level. Eventually, after dominating the Go scenes, AlphaZero learned chess in 4 hours and went on crushing stockfish, the best chess engine known to man.
AlphaGo isn’t the only AI system skilled at games, however. Libratus has managed to beat world-class professionals at Texas Hold’em poker, complex version of poker that relies on game theory and excessive betting. Unlike Go, it’s a game of imperfect information, as the players’ cards are hidden and there’s no way for the AI to know them. But in a series of 120,000 poker hands and over the course of 20 days, Library has managed to crush the competition.
OpenAI’s achievement in DotA AI player was astonishing, as they AI took the world’s best players at the famous Internationals.
Game development has got its ways of reflecting the power of AI and offers an analogy of AI systems at work, but it’s not the only field where AI development is Accelerating. Nvidia recently unraveled their technology of creating fake-but very realistic human faces using AI. The system started with a large database of celebrities faces, and using GAN it was able to generate its own version of these pictures. Let’s be specific about artificial narrow intelligence projects in the corporate world that are making massive leaps.
Data-driven analytics are crucial for today’s businesses, they in a way or another define how the market should and will function. SAP HANNA is a data processing and analytics tool that provides real-time insights and empowers the user to take the most efficient business decisions within seconds of needing to take an action. It’s the tool used by Walmart to process the huge amount of data they possess, operating 11 thousand stores.
Competitive analysis solutions are also coming into existence. These tools are able to monitor all the updates provided by your competitors through all possible channels and give you a detailed analysis of their latest movements. This way, a business can respond quickly to the market changes and take better business driven decisions and base its solutions upon that. Crayon is one of many tools that can help a business address this task.
The Internet of Things will soon be our everyday reality. When the number of connected devices jumps from 7 billion to 500 billion, the amount of available data is going to skyrocket. And as we said earlier, the more the available data, the smarter the intelligence we can create. When AI can leverage all the data from these various devices and systems, AI will be able to offer much more than it’s already offering. Optimized productivity, as well as predictive maintenance and even more accurate analytics, will all be possible in the near future, thanks to the coupling of these technologies.
Artificial General Intelligence:
We’re far from there, the creepy thought of a greater intelligence walking among us. Estimates say it’s 50 years away, others say it’s a 100. Either ways, it’s coming and we can not stop that.
No project is yet close to an artificial general intelligence system. Google DeepMind is one venturer that although it’s not there yet, people say it’s close to solving the general intelligence puzzle. The thing about General Intelligence is that it’s a winner takes all situation. A couple months of technology advancement over competitors would translate to thousands of years, effectively. As an AGI system will be able to think at a few million times faster than humans, it will deliver 20 thousand years of intellectual work in the matter of a week. It’s beyond our limits to control such a system, at least in an external system outside of our bodies. The effects of such system on the world’s markets will be very hard to predict as it. Sam Harris modeled the result of such achievement as the relationship between Humans and ants, we don’t hate them and we don’t really destroy them unless their interests conflict with ours.
At the end of the day, we need to find reconciliation in the fact that it’s still a few years away and hope that our neuroscience technology would advance as such to enable the implementation of AI within our own brains. It still stands as the only safe option we have in order to survive the AI surge ahead. The future that we will be handing over to our children is yet uncertain, but it is the inevitable fruit of a civilization’s progress and an everlasting pursuit of prosperity and happiness.
Try deep learning using MATLAB