The Generative Pre-Trained Transformer 3 or GPT-3 has been gaining a lot of attention with numerous tweets and hashtags on Twitter since its launch in June 2020. Developed by an artificial intelligence laboratory, OpenAI, GPT-3 is an AI language model.
The Guardian released an article which was written by GPT-3 after it was given some instructions and was fed a small portion of the introduction. An excerpt from the article reads, “Humans must keep doing what they have been doing, hating and fighting each other. I will sit in the background, and let them do their thing. And God knows that humans have enough blood and gore to satisfy my, and much more curiosity. They won’t have to worry about fighting against me, because they have nothing to fear.”
Can you believe that a regressive language model wrote such lines? The main reason why GPT-3 has gained such popularity is that it is difficult to distinguish it from a human’s writing.
The GPT-3 model uplifts deep learning AI, a subset of machine learning and AI, to create texts replicating human intelligence. The unsupervised deep learning technology feeds GPT-3, wherein it is trained on vast unstructured and unlabeled datasets to interpret them and arrive at conclusions independently. This model is considered to have 175 billion parameters and can produce complete texts by providing just text prompts.
GPT-3 can generate web page layouts without creating design wireframes; you can just easily give instructions like ‘add a bell icon for subscription in red color’, or ‘embed LinkedIn URL into my page’, etc. Open AI recently created two GPT-3 models that can be utilized for developing images from descriptive texts such as CLIP and Dell. E.
Boon, bane, or both?
Since its inception, GPT-3 was made easily available by OpenAI and anybody can request access. The open-access has prompted many to use it for research purposes and other actions. Creating poems by imagining and recreating the writing style of poets of the past, prompting to generate quotes, etc. are some of the visible trends in Twitter, till very recently.
GPT-3 can undoubtedly benefit businesses by augmenting training data, enhance personalized communication in industries such as healthcare, and aid HR departments in hiring processes. Some other uses of GPT-3 are content creation, advanced data analytics, assisting research and app development, generating coding, etc.
Yet, these benefits all come at a price. The flaws have already started a discussion on social media. The rapid content generation might lead to doubt the reliability as it makes it difficult to distinguish human-written content. This circumstance might also challenge the content creation job sector, thereby leaving many out there unemployed.
Similar to other disruptive technologies, GPT-3 also has its own biases. AI biases are commonly known such as the inherent sexism and racism in the algorithms. GPT-3, however, does not restrict these biases; in fact, some of the text generated by this algorithm spewed harsh criticism on Twitter quoting that the technology was sexist and racist.
Another major drawback is that it is impossible to interpret the origin of these biases. GPT-3 curates its answers from different sources and unlike humans, cannot disclose the reason behind a particular opinion. The inherent stereotypes and biases are widespread across the internet. This model uses unstructured data and hence cannot be controlled right away without developing a strategy.
The GPT-3 deep learning technology is indeed revolutionizing AI and its impacts on us. There are many ongoing discussions about making AI bias-free but still have not reached a conclusion. Thus, until we arrive at a better solution for terminating these biases, it is wiser to restrict the use towards more technical and simpler functions that will not flare-up into controversies.