AI and Your Money: What Are the Issues with Using Artificial Intelligence to Make Financial Decisions?

It is hard to read the news without coming across an article about artificial intelligence (AI). It is very powerful technology with substantial applications and various pros and cons. But, can it help you make better financial decisions? What are the pitfalls of using AI to manage your money?

Let’s take a high level exploration:

What is Artificial Intelligence?

AI refers to systems that simulate tasks associated with human intelligence. These tasks include reasoning, problem-solving, learning, perception, language understanding, and decision making.

ChatGPT is the most well known application of AI right now, but there are other versions as well. ChatGPT is an AI-powered conversational model, also known as a large language model (LLM), designed to understand and generate human language. It learns from the vast amount of text data it’s exposed to during training, and this learning allows it to generate coherent and contextually appropriate text in response to prompts. GhatGPT and other LLMs can understand and generate human-like responses, providing very detailed answers to questions in a conversational way.

How Would Artificial Intelligence Help You Figure Out When to Retire?

So, I asked Chat GPT how it might approach answering the question, “when can I retire?”

It gave me the following approach (which is not actually very different from how you would solve the question using the Boldin Retirement Planner):

1. Data Gathering: AI would collect relevant data about the individual’s financial situation, including assets, liabilities, income, expenses, investments, and retirement goals.
2. Personalization: AI would analyze the data to create a personalized profile of the individual’s financial situation and goals. This would include factors like age, expected lifespan, desired retirement lifestyle, risk tolerance, and more.
3. Market Analysis: AI could analyze historical market data, economic indicators, and trends to assess potential scenarios for investment returns, inflation, and other relevant financial variables.
4. Monte Carlo Simulation: AI might use Monte Carlo simulations to model thousands of possible future scenarios based on various assumptions, considering factors like market volatility, economic changes, and potential investment outcomes.
5. Risk Assessment: AI would evaluate the level of risk associated with different retirement timing options and provide insights into how these choices could impact the likelihood of achieving retirement goals.
6. Behavioral Analysis: AI might consider behavioral factors such as spending patterns, savings habits, and risk aversion to provide insights into how the individual’s financial decisions could affect retirement timing.
7. Sensitivity Analysis: AI could perform sensitivity analysis to assess how changes in variables, such as investment returns or inflation rates, could impact the recommended retirement timing.
8. Real-time Updates: AI might continually monitor market conditions and economic indicators, providing real-time updates and adjustments to retirement timing recommendations based on changing circumstances.

This is a sound process, but the reality is that ChatGPT doesn’t ask questions, it provides answers. And, getting good financial answers requires a lot of very important questions.

When you ask ChatGPT a simple financial question, it will provide a generalized answer. But, it doesn’t usually help you get to a personalized and specific value.

The Difference Between AI and Sophisticated Financial Models for Answering Financial Questions

There are so many different ways to make a financial decision. AI is one option and it will get better. Using a comprehensive financial model is another approach that may be more reliable as of now.

Using a tool like the Boldin Retirement Planner gives you control, accuracy, customization, and consistency. It is the most comprehensive planning model available online and currently has huge advantages over AI because it has much more insight into a user’s financial situation and goals. The tool can provide more nuanced and accurate forecasts than AI can currently. 

The core of Boldin is FPE our Financial Planning Engine. It is the model through which all calculations go through. This model is used every day by an active community of tens of thousands of end users who are continually providing feedback on accuracy and ways it can be improved.  We built and know how the model works and why it works a certain way. The same is not true of LLMs like ChatGPT today.  With LLMs we can’t explain the why or how it comes up with answers, which needs to get addressed before people start making decisions about their life savings on the platform.

A hybrid approach: Financial models currently provide more reliable and defensible answers than AI. However, AI (or human generated articles or consultation) is ideal to assist in interpreting those answers.

We at Boldin are actively working on applying big data and machine learning to plans in addition to considering how AI technology like ChatGPT can be leveraged to help improve education and use of our tools.

What is Missing from the AI Approach to Financial Planning?

The AI approach is sound, but there are several issues and challenges related to the use of AI in financial planning.

1. You have to know the right question to ask: AI is fast and it can be incredibly thorough in answering questions. However, you have to know the right questions (and follow up questions) to ask.
2. Lack of Personalization: While AI can automate certain aspects of financial planning, it may or may not share your values. In planning, there really are no right or wrong decisions, there are only “right for you” choices. For example, you may prefer to mitigate risk and guarantee income while someone else is trying to maximize growth. Your unique circumstances and preferences can be difficult to communicate in AI. All of the non-financial factors that a human advisor and/or you yourself know and should consider are as important as the financial details.

1. Data Privacy and Security: Financial planning involves handling sensitive personal and financial data. The use of AI raises concerns about data privacy and security. Ensuring that data is properly protected and used ethically is a critical challenge.
2. Bias and Fairness: AI algorithms can inherit biases present in the data they are trained on, leading to biased outcomes in financial planning. This could result in unequal access to financial services or biased investment recommendations, potentially perpetuating existing inequalities.
3. Transparency and Interpretability: Many AI algorithms, especially deep learning models, are often considered “black boxes,” making it difficult to understand how they arrive at their decisions. In financial planning, interpretability is crucial for ensuring clients can trust and understand the recommendations being made.
4. Regulatory Compliance: The financial industry is heavily regulated to ensure consumer protection and market stability. Integrating AI into financial planning must comply with these regulations, which can be complex and vary by jurisdiction.
5. Lack of Human Judgment: While AI can analyze vast amounts of data quickly, it may lack the human judgment and emotional understanding that are often crucial in financial planning. Complex life circumstances, changing personal goals, and other factors might require human intervention.
6. Mistakes (Model Reliability): AI models are not infallible and can make mistakes. Overreliance on AI-driven financial planning could expose individuals to unnecessary risks if the models fail to anticipate certain scenarios. A recent study showed that ChatGPT got software questions wrong half the time.
7. Unforeseen Market Dynamics: AI models are often trained on historical data, and they might not be well-equipped to handle unprecedented events or sudden changes in the financial markets, as demonstrated by the challenges posed by the COVID-19 pandemic.
8. Skill Gap: Integrating AI into financial planning requires expertise in both finance and AI. There might be a shortage of professionals who possess both sets of skills, which could impact the effective implementation and management of AI systems.
9. Loss of Human Interaction: Financial planning often involves a personal relationship between clients and financial advisors. Over reliance on AI could reduce the human touch and emotional connection that clients value.
10. Market Manipulation and Fraud: The use of AI in financial planning could potentially be exploited for market manipulation or fraudulent activities if not properly monitored and regulated.

What is the Future of AI in Financial Planning at Boldin?

A lot of the discussion around AI in financial planning revolves around whether or not a human advisor is a necessity or not. Human advisors provide a huge value to many. However, they are expensive and there are not enough of them to serve all people.

We at Boldin believe that the most important human when it comes to financial planning is YOU – not the advisor. We believe that guided financial planning experiences that involve both AI and sophisticated financial models are the future of advice.

A guided hybrid experience using calculations, AI, and human support will insure that your goals, values and priorities are accounted for. Our Financial Planning Model (FPE) will continue to provide mathematically rational answers based on our evolving algorithm. AI will help generate additional options for you to consider and provide interpretation and education about your options. And, we will provide access to human experts when you need the support.

AI will be used to help us continue to deliver on our mission to make high quality and unbiased guidance available, affordable, and understandable to everyone. But, we will never lose site of your humanity and the need for connection and guidance.