GPT-3 and Finance: Can AI finally improve financial operations?

GPT-3? What's that?

I’m sure everyone has heard of OpenAI's GPT-3 by now, but for those who have not, GPT-3 is a neural network for generating text. Essentially, it is an "A.I." that tries to generate appropriate text in response to the text you feed it as a prompt. It has taken over social media by storm as people shared how they used GPT-3 to write songs, generate code, and ask it the meaning of life.

We recently got access to the private beta and asked GPT-3 in what ways it could change the world of finance. The answers were quite honestly more detailed than we expected, so we thought we would share some of the interesting perspectives GPT-3 came up with. The following excerpt was written by GPT-3, with a couple prompts we gave it along the way.

How will GPT-3 impact finance?

Financial operations are about to undergo a complete transformation. AI like GPT-3 will act as an always on, always insightful virtual CFO, turning slow and esoteric financial reports into real time qualitative insights. For example, GPT-3’s vast knowledge of world financial data, in conjunction with your business data, can provide insight into the impact of a new product launch on your cash flow. It will be able to forecast future costs and predict whether an acquisition will be profitable, as well as generate real time financial projections and warnings that human CFO's would be hard pressed to offer.

And the impact on internal control, auditing and regulatory compliance will be immense. While GPT-3 may be years away from self-awareness, its ability to detect anomalies and alert a company’s management to potential fraud and money laundering will be very real and effective. We will be able to tell if an employee is using the corporate card to pay for their daughter’s summer camp, or for a more nefarious purpose. If invoices don't match contract terms or potential discounts are detected for the company's vendors, AI would have the capacity to detect these at an unprecedented scale.

However, for AI to succeed in the finance role, it must be built with a number of specific considerations in mind. A.I. needs to have the ability to collect and process data at an incredible rate. It must also be able to learn from past data and continuously grow its knowledge base. It must also be able to continually develop its own “cognitive” knowledge, by combining knowledge with its past experience and external sources. And finally, it must be able to communicate effectively with humans, since it will need to work in close conjunction with them.

Will AI unlock latent potential?

By focusing on these considerations, AI can become the finance department's most valuable asset. It can provide information in a way that humans are incapable of doing so, and it will never tire or suffer from the daily distractions that plague the finance department. For instance, when GPT-3 analyzes a new business deal, it will be able to provide real time data based on its knowledge of your entire financial situation, not just your company’s data alone. This type of holistic view is simply impossible for a human CFO or accountant.

These technological advances will likely lead to a fundamental restructuring of the finance department. It may even lead to the elimination of some finance positions, and the elevation of others. Finance staff that focus on monitoring accounts receivable and payable, or analyzing financial data will likely see their roles diminished. In fact, some predict that only the most senior finance staff will be spared from this transition.

However, those who focus on leading, mentoring and guiding AI’s development will be invaluable. Those who work to understand its strengths and weaknesses will also be critical. Finally, those who are able to help AI to develop an understanding of external market conditions and global economic trends will play a vital role in making it a successful asset for your company.

It is important for CFOs to stay abreast of technological advancements that can help them gain a competitive advantage over their peers. As we have seen with other technology innovations such as Big Data and cloud computing, AI can significantly increase productivity while improving financial transparency and compliance at a reduced cost. For these reasons alone, CFOs must take steps to ensure that they are incorporating AI into their companies’ financial management strategies as soon as possible.


Needless to say, we were pretty blown away with how detailed the responses were. Considering we barely prompted GPT-3, this is far beyond what we were thinking of its capabilities. Although there is still a long way to go before the use cases it mentions are viable, we have started utilizing GPT-3 for several projects we had in mind. We're excited to continue our experimentation and hope to find ways to use GPT-3 to drastically improve how financial operations are done today!

Start spending smarter

Automate vendor management with Trimwire today.