One cannot enter a room of professionals in any industry right now without the conversation drifting to artificial intelligence (“AI”). Every mixer inevitably includes the question, “how are you using AI?” The responses range from “we are dabbling in AI search/chat,” to “we have reinvented every process around AI and trained a fleet of bots to do the work; let me introduce you to our cloned CEO bot.”

My own reactions to these conversations range from curiosity and hope to concern. Often the concern hinges on my belief that there is something sacred to being human, which could be in jeopardy.

My father-in-law showed me a video of a puppy taking care of a baby. He unwittingly commented, “how cute is that?” to which I replied, “it’s cute, but it is AI, see how the fur changes when the puppy moves?” His face showed a familiar disappointment, and I told him about being similarly disappointed after feeling moved by a poem, only later to find out it was written by AI. Why do we experience this feeling of letdown? Because the source matters! At the end of the day, we are unconsciously seeking human connection. When I read that poem, I thought I was connecting to a human’s words, ideas, experience, and creativity. When the poem was revealed to be AI-generated, I felt a loss of connection and, ultimately, distrust. For this reason, the words of each Soundings post spill onto the screens from a human brain. Our readers want connection and perspective, which we will deliver.

Before drafting and implementing TCR’s AI policy, I asked myself: what is uniquely human about our work as a bespoke OCIO for multi-billion-dollar families?

Relationships
A bot explaining a downturn in a market will never be as effective as a human. As humans, we are exposed to risk and experience emotional responses such as fear, anxiety, uncertainty, and worry. AI telling me not to worry cannot ease my mind, no matter how logically or articulately it approaches the subject. I cannot shake AI’s hand, look it in the eye, or build a human connection. What is needed is an experienced, trusted advisor who has built rapport, who understands both the logical side of investing and the human experience. AI can imitate empathy, but it cannot feel it, and that is an insurmountable barrier to trust and connection.

Discernment
Anyone who has spent enough time using an LLM has an example of AI getting it wrong. I once was working with AI on an Excel formula to calculate RMDs considering tax code, inheritance types, claiming elections, ages, life expectancy tables, and dates. Several times, AI produced a formula misinterpreting the logic behind which life expectancy table to pull from. I replied in all caps about its error and put the bot back to work until it produced an accurate formula. How did I know it was wrong? After working on client accounts for 12 years, I had very intimate knowledge of RMD calculations. For this reason, we cannot ask AI to do something beyond our own ability to review and validate the results.

Decision-Making and Accountability
With bots, there is no literal or figurative skin in the game. If an error is made, AI bots are not calling the client to apologize, own up to their mistakes, or set the record straight. AI cannot feel shame or embarrassment. With humans, every interaction we have is intertwined with our reputation, relationships, identity, and at times, our livelihood. AI cannot serve as a fiduciary; that responsibility is uniquely human. This brings me back to a presentation by Cary Weston, podcast host of the ChatGPT Experiment. He stated that after growing tired of his AI Agent’s errors, he needed his bot to understand the importance of getting it right. He told his bot that if it kept making mistakes, the results were going to be poor, and if the results were poor, the clients would be dissatisfied, and if the clients were dissatisfied, he would lose business, and if he lost business, his wife and children would starve. This did not “teach” the AI anything; it only revealed that AI has never experienced hunger.

Governance
Governance comes from the Greek word “kubernan” which means “to steer” or “to pilot” a ship. In practice, governance starts with getting the right people in the room, then determining the proper course in alignment with mission, vision and values, and lastly tracking and modifying it over time. Doing this well requires relationships, discernment, and accountability, all topics we previously determined are not suited for AI.

A family is not a single person with a single goal. It’s usually a complex, multi-generational, multi-entity group seeking clarity and a path forward. Families rely on a carefully curated group of trusted advisors, not only with expertise in their fields, but also with a deep understanding of the family and its goals. Proper governance pulls the family together by bridging generational tensions and competing interests to form a collective agreement around purpose, values, and impact.

A high degree of discernment is needed to arrive at a tailored and enduring Investment Policy Statement (IPS) for a family. No two families are the same and for that reason, the solutions should be unique. The IPS is, in essence, the setting of the course a ship follows. AI can produce an IPS, but it would be cookie-cutter and ineffective without a human with emotional intelligence pressure-testing it. It might lead in a direction unintended or require constant course corrections. AI would not be accountable for these errors of judgment.

The proper course for a family can evolve over time. There are external forces, like changes in the markets, interest rates, a liquidity event, inflation, and threats and opportunities that arise and resolve. There are internal forces like marriages and divorces, births and deaths, next generations coming of age and previous generations taking a step back. Those at the table need to continuously notice these internal and external changes, ask the right questions, and adjust the course as needed. A fiduciary investment adviser is distinguished not merely by the quality of its analysis, but by its consistency, and its accountability for the consequences of its advice. It takes people and relationships built on trust to steer.

So, considering all of this, how is TCR using AI? We are keeping humans at the helm for final review and all that requires connection, discernment, and accountability. We are engaging AI for what is uniquely artificial, which is where we will pick up in next week’s Soundings, Dopamine Donuts.

Back to Soundings

Tide Cycle Resources (Tide Cycle) is an investment advisor registered with the U.S. Securities and Exchange Commission. Registration does not imply a certain level of skill or training. A copy of Tide Cycle’s Forms ADV Part 2 and Form CRS are available without charge upon request. The opinions expressed are those of Tide Cycle. The opinions referenced are as of the date of publication and are subject to change due to changes in the market or economic conditions and may not necessarily come to pass. Forward-looking statements cannot be guaranteed. Nothing contained in this document may be relied upon as a guarantee, promise, assurance, or representation as to the future. This should not be taken as specific investment advice. We recommend consulting an investment/tax professional before making financial decisions based on any information provided.

Not all posts are archived here. Please reach out if there are other topics that interest you.