We have all seen sci-fi movies in which an Artificial Intelligence (AI) can communicate with humans in a way that is indistinguishable from real humans.
A task that has proven to be very difficult in real life. Languages don’t have fixed rules, are ever-changing, and are not set. While humans have an intuition to pick up a language, it is very difficult to teach this to a machine.
In 1950 Alan Turing proposed a test to determine if a computer can display intelligence that is equal to that of a human. If artificial intelligence responses are indistinguishable from a human, it was deemed to have human-like intelligence. Today more than 70 years later we can confidently say that computers are able to pass the Turing test, as we will see in this post.
Although still very debatable if we can call this true intelligence because AI is usually limited to a specific or narrow area. It lacks the combined abilities to reason, solve puzzles, make judgments, plan, learn, and communicate. It does not have the capacity to have objective thoughts, consciousness, and self-awareness. An AI having these capabilities is a theoretical concept which we call ‘True Intelligence’ or ‘Artificial General Intelligence’. This type of intelligence is not in the scope of this article as it will be focusing on the specific area of natural language understanding. In recent times big progress has been made in the field and we can expect a surge of applications that will change our lives.
One of the most exciting fields in AI: Natural Language Processing (NLP) handles tasks like understanding language. There already exist a wide variety of applications that make use of NLP like translation, voice assistants, and chatbots. We will have a look at some amazing new possibilities of NLP.
The term Natural Language Processing (NLP) was first coined by Stuart C. Shapiro in 1964. His definition of NLP is: ‘the area that studies the nature of language and develops techniques for processing natural language data’. In the following years, many different approaches to NLP were developed. A lot of these approaches were based on the symbolic approach to AI. In the beginning, NLP was used more as a tool to solve other problems in AI. One of the first applications of NLP was the development of programs that could translate between different languages.
Data in the form of text is getting more and more common. Businesses have plenty of text-based surveys and emails to go through. Researchers often use social media posts for analysis. So, it should be no surprise that NLP is becoming a must-have skill for data scientists. Companies like Google and OpenAI have been making big progress in the field of NLP. Models are trained to replicate and learn how language is used instead of the more traditional pre-programmed rule-based approaches. An example of this is GPT-3 from OpenAI is a model that has been in development and is able to generate and autocomplete text based on prompts from a human writer. Google LaMDA is a similar model that is trained on how human dialog is performed.
Next, we will discuss some amazing and surprising applications that these new technologies can do.
In the following example, a summary of the text is made in very easily understandable language. Simply by asking the AI to make the text understandable for a 2nd grader.
By listing a recipe and ingredients the AI can give you the instructions on how to prepare it.
In the case of writer’s block. The AI can help you give an outline for an essay.
Don’t know how to code? Just ask GPT-3 to build an app for you by describing it.
Todo App:
Recreate Google home page
This AI helps you get a date. It can create your dating profile bio and can give an original opener.
As you can see the possibilities are endless. As an experiment we added a paragraph in this article that was entirely generated by AI, can you spot which one it is????? Contact us via LinkedIn or our website if you’re interested in applying this amazing technology in your organization or when you have any questions regarding this post.
Back to blogs