Artificial intelligence has dominated the spotlight for the past several months with its headline-grabbing abilities to generate text and images — and disrupt markets.
While it’s difficult to separate hype from fact, there’s little question that, for forward-thinking small and medium-sized businesses, AI and AI-driven automation can unlock game-changing breakthroughs in productivity, cost savings, and competitive advantage.
Already, we’re seeing the first wave of these advantages as finance departments start to harness AI to streamline their tactical operations. They’re recognizing and categorizing financial data to automate accounts payable processes to reduce errors and increase productivity. They’re using AI to review the general ledger to identify patterns in transactions and detect anomalies that might elude human review and that merit further investigation to prevent fraud and ensure strict compliance.
AI today and tomorrow
Even generative AI tools like ChatGPT, Google Bard and recent updates to Bing can play a role in finance. For instance, they can summarize new changes to tax codes in different jurisdictions. Or you can use generative AI to answer customer emails inquiring about invoices, payment or credit terms (with appropriate restrictions and protections for sensitive information).
Adopting these (and similar) uses of AI will make finance more efficient, but more importantly, they will also separate winners from losers. In my opinion, it isn’t hyperbole to say the meteoric emergence of AI is not unlike the arrival of the internet or the smartphone. The only key difference: AI’s market impact will be felt even faster and wider.
AI will transform the rapid exchange and understanding of accounting and financial information. The classic “data-in, analysis, data-out” cycle will be much easier and faster for the accounting and finance team who will gain a clearer picture into the health of the business. We can expect that soon, AI systems will be coupled with deterministic systems to form a complementary solution that provides the best of both. Generative AI solutions will become data- and system-aware, determining whether the answer can be generated or if it requires another system for accurate inferences. We also expect to see human-friendly interfaces interacting with complex software systems that summarize the output and present it in simple, digestible formats.
Note that the power of AI isn’t limited to its ability to improve efficiency and productivity for finance and accounting. While it excels at automating repetitive tasks and extracting insights from large volumes of data, it also offers the ability to understand patterns in historical data and predict future recurrences. With its ability to detect trends, AI allows professionals to identify opportunities and risks that may have gone unnoticed. When AI-generated insights are combined with the expertise of finance professionals, they can lead to better decisions, better performance, and ultimately, better outcomes.
Risks and limitations of AI in finance
Let’s also be clear that, concurrent with its promises, AI also presents numerous pitfalls that must be understood and avoided. If we trust AI too much and create processes that lack human intervention or involvement, we face a significant risk of error that may or may not be detected. Hands-off AI can be a recipe for disaster, so we must design-in human checkpoints, audits, inspections and quality reviews.
AI can lead you to strategic decision points quickly — by performing research quickly and uncovering compelling insights in financial data sets. However, it’s unlikely you want to rely on AI to make strategic decisions. For instance, asking an AI tool, “Where should I invest our money to yield the largest possible return?” might place too much trust in a still-evolving technology. Instead, you should use AI as a superb starting point, but it shouldn’t make all the decisions for you. Humans still think outside the box and take advantage of novel circumstances — and AI can miss those opportunities.
We must also factor in the technical limitations that are clearly present in today’s AI technologies. Through its so-called “hallucinations,” generative AI solutions can generate highly convincing and highly erroneous results. Of course, despite all of the hype, we are only in the very early phases of the AI adoption cycle. We can expect rapid and frequent improvements to the technology that will address these flaws. But tools can also enhance and accelerate these improvements with feedback loops to notify the AI tool if its inferences, analyses and predictions were accurate. That helps AI models get better over time.
Ethics of AI
When you’re considering how you want to implement AI in your organization, it’s important to initially select projects that — first and foremost — “do the right thing.” For instance, deploy AI tools and solutions that won’t perpetuate bias. That might mean you train your lending applications using data sets that are free from bias about application approvals. Or you can sift through your previous lending applications to search for patterns that indicate systemic flaws in the lending process.
AI will raise the bar for the work that finance does. Instead of replacing human jobs, it will create a whole set of possibilities. However, we should be clear in understanding that AI is no panacea for finance. Judgment, experience and guidance — along with trust, security and ethics — are non-negotiable requirements. These will only accrue gradually over time. Forward-thinking finance pros should recognize that AI is here to stay, it will improve, it will offer benefits and risks, and it will eventually transform how we think about the operation of accounting and finance.
Credit: Source link
