Tech billionaire Elon Musk was always pro-technology. But way back 2014, he warned us of the dangers of AI . He even says it is a greater threat to humanity worse than nukes. Is he a techno prophet warning of a ‘Terminator’ apocalypse?
What is AI?
The Encyclopedia Britannica defined Artificial intelligence (AI) is the ability of a computer or a robot controlled by a computer to do tasks that are usually done by humans because they require human intelligence and discernment. Although there are no AIs that can perform the wide variety of tasks an ordinary human can do, some AIs can match humans in specific tasks. AI enable computers to talk and speak like human (natural language processing) and see the world as humans see (computer vision).
Machine Learning
Artificial Intelligence can diagnose diseases by looking at photographs of tissues or carry out task by the voice command. Narrow AI can perform amazing but only specific task. Machine learning is a field in narrow AI that powers computers to learn by mining thru large amount of data, looking for patterns and make predictions based on the pattern. A neural network is computer hardware that makes machine learning possible. The neural network mimics how the brain works. It process information like the brain by sending a neuron (node) an electrical signal to other neurons and return feedback. Examples are voice command for Siri and Alexa and facial recognition in security cameras and Facebook
Elon Musk pointed out that narrow AI is a not a species threat to humanity. However it is a disruptive technolgy that can result in the loss of employment and even wiping out entire industries
On the other, super Intelligent AI allow computers to think like humans. Super Intelligence is the holy grail of AI and many companies are racing to be the first. One approach is to join several narrow AI to work together to form a more intelligent AI.
Image:
Google did just that by putting an AI brain into a robot body. It’s PaLM-SayCan, an AI that can understand natural human language. And World Grounding which It gives the robot the ability to decide if a task is physically possible. Computer vision enable the robot to navigate the real world. In a demo, the human operator tells the robot to fetch the red cup. PaLM-SayCan AI interprets verbal command then sends the task to the World Grounding who in turn send low level command to the robot’s arm if possible. Computer vision locates the red cup and robot completes the task. Google experiments shown that the robot completes the task 74% of the time.
Advances in technolgy will make AI learn at an exponential rate.
Following Moore’s law, computer hardware will have the computing capacity of the human brain by the 2030s. Some experts believe the software that will power this hardware would be develop by 2040. By then, Singularity, the point where the computer has surpass human being has been achieved.
This is the danger Elon Musk feared. He warned of two possible scenarios.
- A evil dictator can seize control of the AI and bring about the next ‘Dark Ages’.
- AI will rise up and destroy all humanity and bring the apocalypse
Either scenarios is grim, but Elon Musk has a plan.
To be continued at part two.
