how to use the three types of artificial intelligence

your firm will need to redefine its processes.

by hitendra patil
rise of the aiccountants

artificial intelligence (ai) is a rage worldwide now, and accountants are racing ahead to test/try out publicly available ai tools. ai-powered tools can potentially automate tasks outside the purview of the software solutions created by programmers. ai also has the potential to help accountants to better understand and analyze data, which is essential for helping their clients, and accountants themselves make better informed business decisions.

more: ai means clients need you more than ever | seven changes ai brings to accounting | humans beat ai in some aspects | why accountants should embrace artificial intelligence | ai brings opportunities for accountants | seven things you must know about ai
exclusively for pro members. log in here or 2022世界杯足球排名 today.

all of you (almost) have already tried chatgpt. but it is just one type of ai – a generative, large language model (llm). as it is “language”-based, it can, right now, quickly help accountants enhance marketing and client communications.

but what about the core accounting, tax and payroll work?

while i was researching, it became evident that there will be three main ways in which ai will shape itself in the accounting profession in the future: public ai, private ai and custom ai.

public ai, such as chatgpt, will be available to anyone.

• it can typically be used for basic tasks such as simple data analysis, classification of transactions, helping make excel usage faster and easier, etc.
• public ai is often less expensive for a user than private ai or custom ai, but it is also less customizable and may not be as accurate.
• anything “public” in the ai world will level the playing field, and even accountants’ clients can use public ai tools to generate information and insights accountants now create.
• the challenge for clients will be the relatively non-trustworthy accuracy. but, even then, the common “information exchange” demands on accountants are likely to reduce because of public ai tools.
• if people make business/tax decisions based on public ai, it is likely that “resolution” work will only increase for accountants.
• given the responsibility of accountants to protect clients’ confidential data, data privacy and security concerns will remain, even with more investments by public ai creators to address these concerns.

private ai, such as bloomberggpt, will be developed specifically by/for some particular company or organization.

• private ai will likely be more expensive to build than public ai but will be more industry/profession-relevant, accurate and customizable.
• in the accounting profession, private ai tools will most likely be built by the technology vendors that provide solutions to accountants. cloud solutions companies with access to the data of hundreds of thousands of companies, millions of individuals, and billions of transactions are sitting on a gold mine of foundation for private ai.
• private ai can provide a significant competitive advantage for accountants who use it effectively.
• there will likely be hybrid ai tools that are a fusion between private and custom ai.

custom ai will be developed specifically for a particular task or set of tasks or for/by a given firm.

• such custom ai tools may get developed within the private ai frameworks. imagine big four or top 100 firms building ai tools based on their client data.
• custom ai will likely be the most expensive type of ai. but it will also be the most customizable and provide the best competitive advantage of the intellectual property rights of the firm.
• it is very likely that “ai platform as a service” – aipaas – will enable firms to set up their own cloud infrastructure to create custom ai tools with their own data. imagine your firm being able to build custom ai based on your firm’s clients’ data.
• for smaller (even midsized) firms, the “training data” available may be inadequate to become comprehensively useful and competitively advantageous.
• firms will be able to use custom ai to develop new and innovative solutions and services.
• there are already cloud-based ai tools that can help you create chatbots based on your data (such “training data” will have to be uploaded to the “cloud,” though, thereby causing some privacy and data security concerns).

what does it mean for you?

the accounting profession is still in the early stages of adopting new-generation ai. however, ai will rapidly become a critical tool for accounting firms.

as ai progresses and makes more inroads into the day-to-day technology solutions accounting firms use, firms will need to redefine each of their processes. in addition, it will need more and better checks and balances to ensure ai-processed work is reviewed for accuracy, to ensure biases haven’t crept in, and for explainability.

as ai becomes increasingly ingrained in your work and processes, you can explore the possibilities of making the most of private ai included in your software solutions and even creating custom ai tools for internal and client-facing processes.

getting familiar with the capabilities, advantages, limitations, concerns and potential pitfalls of ai systems is the first step toward becoming an aiccountant, i.e., an accountant augmented by ai. however, your aim in using ai is not just about “using ai,” but it is about using ai to deliver better, more impactful services to your clients, making your firm more efficient, and creating work processes that unlock the human potential of your staff.

posted on february 4, 2024