Posted September 17, 2018 06:14:15The next step in the artificial intelligent industry, which is already well on its way, is to turn the machines on and off, said James Anderson, CEO of AI-focused venture capital firm Andreessen Horowitz, in a podcast interview with the New York Times.
“We’re not going to go out and say, ‘We’re going to turn on and stop everything,'” he said.
“But if there is a big breakthrough in artificial intelligence, we will do so, and we will have to make it happen.”
While the technology may seem complex, the process is relatively simple.
A computer, or AI, creates an image of a specific object, like a photo or a house, then uses that image to analyze the object and identify it as a possible target for its AI.
The process of using an image to identify a target, or “tagging,” is similar to the process of finding a scent in the environment.
A dog might find a dog that smells like a cat.
The image is then used to train an algorithm to identify the object based on the scent.
The algorithm can then determine what a particular object is capable of, such as a gun or a bomb.
It’s a system that Anderson likens to a natural language processing (NLP) system, which uses artificial language to create a conversation between two people.
It’s similar to using a natural speech recognition system, such the ability of a robot to speak.
But, Anderson says, artificial intelligence is just one of many fields in the AI space.
He notes that the real challenge for the AI industry is not just getting AI to learn and recognize a target; it’s also to use the information in the image to build a more intelligent system that can be used for different purposes.
“A big part of what we need is a machine that can think about a task very quickly and very quickly understand the problem and then make that decision about how to proceed,” he said, adding that artificial intelligence companies are “doing it differently.”
“If you’re trying to build something that can do all of these different tasks, you’re going through different stages of thinking, like, ‘What do I need to do to make this better?'”
“So you’re in the beginning stages of learning, then you have a really deep learning process where you can build things up to a very high level of accuracy, but then you’re building these things from the ground up.”
This is similar, Anderson said, to what companies are doing with facial recognition technology, which he describes as “a lot like the natural language process.”
The technology that’s being used today, he said is very similar to what was used by the military in World War II.
“I think you see some of the things that you see from this in the military,” he continued.
“They’re building their weapons systems from the beginning.”
He continued by saying that it is likely that in 20 years, AI-based products will be able to understand language in their own right.
“There’s no doubt that this is going to happen,” he added.
“The way that the technology is going forward is going towards, ‘How do I use this information to build these things?'”