The year 2020 is likely to be remembered for the expect the unexpected, COVID-19 pandemic besides the significant presidential election. There is another new contender for the most spectacularly newsworthy happening of 2020- OpenAI’s new language generator- GPT (Generative Pre-training)-3.
GPT-3 has been developed by OpenAI, an Elon Musk-backed artificial intelligence lab based in San Francisco. GPT-3 API provides a general-purpose “text in, text out” interface, allowing its users to try it on virtually any English language task. GPT-3 is simple for anyone to use but also flexible enough to make machine learning teams more productive. Users can request access in order to integrate the API into their product, develop an entirely new application. GPT-3 is trained on uncategorized internet writings, basically guesses what text will come next from any starting point. The GPT-3 language model is built for guessing with 175 billion parameters which are a mammoth 10 times more than its previous competitors and so surprisingly powerful.
The eventual uses of GPT-3 are tough to ascertain, but it has a huge untapped potential. GPT-3 conversing at a conceptual level can translate language, answer email, perform some programming tasks, help with medical diagnoses and may serve as a therapist in some years. The amazing language generator offers a surprising degree of sophistication and is developed to write poetry, dialogue and stories – a typical failing for many automated response systems. Users can also ask questions about God from the GPT-3 language model.
Though the language generator is built for generality and depth; it comes with its loopholes as well. Canvassing bias, it may answer many queries, at least in its current state. As a general philosophical principle, the GPT-3 accepts that being weird sometimes is a necessary part of being smart. Like so many other technologies, GPT-3 has the potential to rapidly improve, being still at the nascent stage.
GPT-3 is 100x larger than its predecessor GPT-2. It is far more competent than its predecessor due to the number of parameters it is trained on175 billion for GPT-3 versus 1.5 billion for GPT-2. OpenAI has started to release the GPT-3 API it to a select few people last week who had requested access to a private early version, current the private version has a waiting list, which ascertains the popularity of the API.
Addressing Business Dynamics
OpenAI was set up as a non-profit with a $1 billion pledge from a group of founders that included Musk. In February 2018, Musk left the OpenAI board but continues to donate and advise the organization.
OpenAI made a profit in 2019 rising US$1 billion from Microsoft to fund its research. GPT-3 is set to be OpenAI’s first commercial product with Reddit signing up as one of the first customers. GPT-3 points to the health and dynamism of the Bay Area tech world, and of the US economy on a macro level. The innovation has come to the U.S. land before China could reflect the power of decentralized institutions. OpenAI started releasing it to a select few people last week who had requested access to a private early version, and many of them have been blown away.
In the long run, OpenAI wants developers to play with GPT-3 and see what they can achieve with it, before rolling out a commercial version later this year. It is unclear how much it will cost or how the system might be able to benefit businesses, but it holds a significantly high potential takeaway to improve chatbots, design websites, and prescribe medicine.