OpenAI’s Chat GPT blew up in a blink. For obvious reasons, everyone loves this technology. Its interactive approach mimics human intelligence, but given the terabytes of data it is trained on, GPT can surpass human aspects of motivation, creativity, and oversight.
Bloomberg has taken this technology and transformed it for their use. GPT stands for Generative Pretrained Transformer. It is based on NLP, or Natural Language Processing, which allows humans to communicate with computers in a meaningful manner. It is being used by students, working professionals, researchers, and entertainers alike. Given the variety of data that it runs on, its use is extremely versatile.
Bloomberg’s altered approach
Bloomberg saw the potential of this technology in improving their interface. Their internal AI model allows them to communicate better with computers. Using NLP, the AI transforms human text into that which computers find easier to process.
It can also generate headlines based on information that the user types up. Not only does it paraphrase efficiently, but the system is also able to analyze which headlines will perform well and which might not. It can answer questions like “Will this headline appear bearish or bullish to investors?” and “CEO of Citigroup?”.
Unlike OpenAI, Bloomberg’s GPT will be installed in their terminal software and is trained on terabytes of data specific to business and finance. Its outcomes are riveting for the tech industry and can alter the company’s success in the upcoming years. The AI will be able to replace human tasks of creativity and research, leaving employees to take their roles one step further.
Of course, as technology grows, concerns about human jobs are always present.
The AI Market
This does open up the market for AI companies, who recognise their value. Even AI startups are getting funded in billions. Head of ML Product and Research, Gideon Mann, reveals that the performance of GPT-3 in language modeling was state-of-the-art and was an awakening. “It is going to change the way we do NLP here”, he claims.
Bloomberg’s swift and suave move shows how the AI market may not be dominated by giants, and how there is room for everyone to join the movement. By adopting their systems specific to their purpose, Artificial Intelligence can be made even more intelligent.
With big market shares, OpenAI and BigTech seemed intimidating to smaller tech companies. Given their impregnable amounts of data, it would be unavoidable that other companies would have to buy their data to set up their own GPT systems. Bloomberg’s use of free access to off-the-shelf data restored hope for future startups wishing to take GPT to the next level.
With specialized task outcomes like transforming company names to stock tickers, highlighting important names in a given document, and predicting the performance of a headline, Bloomberg’s altered approach sets the tone and inspires other companies to make the most of the technology. Compared to OpenAI’s ‘foundational’ model, Bloomberg’s more tailored approach is more sophisticated and resourceful when a specific goal is to be achieved.