Is creating a self aware AI possible?

“green and red light wallpaper” by Pietro Jeng on Unsplash

Is creating a self aware AI possible?

This is a question that has been pondered by researchers over the past few decades. Can you really create an AI that is self aware like humans. And the answer to this question depends upon what your beliefs about consciousness are.

We will be covering the main believes held by scientists about consciousness in the next few articles. Subscribe to stay tuned. For this article we wanted to test out the most commonly applied belief when trying to create self aware Artificial Intelligence.

A common belief held by computer scientists proposes that consciousness is the cognitive process of using old and new information into perceptions and actions. If this explanation is to believed then machines one day will become the most powerful consciousness as they can process and save vast amounts of data and libraries in milliseconds and logically solve complex problems than any human would possibly be able too.

So we ran an experiment and using a mixture of heuristic pattern matching, recurrent neural networks (LSTM) and knowledge based wolfram programming, we trained our AI with limited dataset to match the best answers from its knowledge base while also generate its own responses. The training data used was movie scripts, books and personal chat logs.

The above video shows the results. As you can see the AI was confused during a few parts, because it is still learning. However, the AI was mostly able to have a nice and interesting conversation with the creator.

Our next step is to find a model that allows AI to learn more efficiently when trying to generate it’s own responses like a human being. Stay tuned for more interesting updates and to learn how to utilise AI for your personal or professional use without coding.

#bot #artificialintelligence #neuralnetworks #selfawareness