This is the story of humanity being forced to choose between being on the right side of history and doing what he loved or earning a living. The choice seemed clear at that moment: when you find yourself under a runaway train, there is no time to think about the consequences of your actions. You just have to let go and start tumbling in its direction.
The term artificial intelligence was first used in 1956 at a Dartmouth conference, where it was defined as “the science and engineering of making intelligent machines”. It has evolved since then to include computers being able to perform tasks normally requiring human intelligence. Tasks commonly associated with AI include speech recognition, computer vision, planning, learning, problem solving (E.g. computer programs that can solve problems that are easy for humans, such as those in sudoku), natural language processing (communication and machine comprehension of human speech and writing) and many others.
The world today is full of algorithms that are making decisions on our behalf; from Google Search results or Facebook newsfeeds, shopping recommendations, movie recommendations, spam filter, to self-driving cars.
The biggest misconception about algorithms is that they are impersonal and objective: in fact, the data going into an algorithm can either be chosen with enough care or be polluted enough to give it a political bias – this was proven recently when Google Photos tagged African Americans as gorillas.
In short AI is the technology that can make our lives more convenient, save us time and money, create new jobs and opportunities that we cannot even imagine today.
Strong AI is a product that exhibits full and unsupervised intelligence. That means it’s capable of doing things like learning, planning, reasoning and acting – all the other forms of intelligence that humans are capable of performing.
Today there are two main strands of research into artificial intelligence – Strong AI (the one we would want to create) which is mostly done by academics, and Weak AI (sometimes called narrow AI) mainly done by big tech companies such as Google , Facebook or IBM . The latest advances in weak ai are based on neural networks (machine learning), deep learning methodologies and input from psychology for understanding human perception.
Strong ai will be achieved with an approach known as Artificial General Intelligence (AGI).
The term was popularized by Vernor Vinge in a 1993 op-ed piece for the ” Los Angeles Times ” titled “Writing Prompt #3”, later reprinted in his essay The Coming Technological Singularity :
At that point, human life will be irreversibly changed. We will be able to instantiate any conceivable mind (living or dead), on any planet in the universe, and we can do it with matter already available in nature. It is time to stop anthropomorphizing the future of artificial intelligence: There is no reason to think an AI would develop a conscience, feel compassion, or follow [any] moral code. The only way to win is not play at all.
Basically, the singularity is a technological event that will happen in the near future, like it has happened before many times in human history and like they could happen again. It is when machines develop their own consciousness (aka strong AI).
The answer to this question is not simple and depends on whom you’re asking. Most people in the field believe that AGI will be achieved within the next few decades. However, over the past century we have seen claim after claim that it is right around the corner only to be disproved – such as the idea that only ten years were left to crack the code and do voice recognition, or even that computers would not be able to beat a grandmaster at chess within the next 10 years.
There are many reasons to believe that AGI will be achieved, some based on the way technology has been developing in the past 50 years, some based on the technological and economic trends of big corporations today and some purely rational. Another point is that if we ever had to rely on AGI to keep us safe from other dangers (such as nuclear bombs or climate change), then that is probably a good enough reason to invest into it.
Computer vision has made huge advances through the use of convolutional neural networks. The first network that could recognize a cat with reasonable accuracy was released in 2016 . Speech recognition systems have also improved remarkably, and mobile phones are able to understand us as well if not better than humans. However, still their understanding is limited – they can’t really have a conversation yet.
Key differences between Weak and Strong AI
The main difference between weak ai and strong ai is that, while the first can be thought of as human-level intelligence in a specific narrow domain such as playing chess, driving cars or recognizing pictures, an AGI would possess human-level intelligence across many domains.
Yes, the bots are here! They can even sign NDAs. Some companies have already started using AI to write content for them. Companies such as Narrative Science (using their technology Quill) or Aylien .
This is good news because it means that there will be no need to pay freelancers anymore and you can hire just one person to do the job of ten.
Why is it important to have an editor?
The answer is simple: because computers can’t think like humans and their writing will always be sterile, boring and devoid of soul – though with a good enough understanding of your business (and your client) they will accelerate your content creation.
Peter Thiel and Elon Musk invested $1 billion into a fund called Open AI to promote and develop artificial intelligence in positive ways. Their strategy for this is to create an open source platform for strong AI, i.e., GPT3 ( General Neural Network ).
Using the same techniques they are testing on Dota players, they want to create a positive artificial intelligence.
so many AI services….
a vast majority of them are powered by GPT-3, a general Artificial Intelligence algorithm that is open-source and trained to create any content.
The GPT3 should be available in the cloud, and allow developers to create their own artificial intelligence engine by training it with their own datasets. AI will be able to rewrite its code using genetic algorithms until they are intelligent enough for your business needs.
Not every platform which is connected to GPT-3 is equal, because each platform will have its own way of training the GPT3 and its own interface.
This means that the success of each AI platform is going to be tied with its user-interface and methods for connection with other systems (like human-AI interfaces).