Hi Reader,

The U.S. government shut down Claude Fable 5. I thought to myself, "That's a bummer. I'm revising my website for the third time. This will make things more difficult."

That was June 12. The directive was an export control order, and it forced Anthropic to disable the model for every customer. Access came back on July 2.

This is my actual view. Yes, 100%. AI is here to stay, and it's not going away. Quick qualification: It's nearly one in five people who went looking for advice last year. The financial advisor industry skews toward people who have assets, high income, wealth, and the ability to pay for financial advice.
​
​

Where It Actually Helps

AI can help in powerful ways. Most of the frontier models from Alphabet, Anthropic, and OpenAI have deep research capabilities. Is it perfect? No. Are humans perfect? No.

In my practice, I use AI for a variety of operational responsibilities. Many people, for example, Claire Vo, would describe an AI agent as an "employee." It can type emails, answer phone calls, draft content, book a business trip. The line between human and agent is blurring.

Human-in-the-Loop

And still.

It's hard to remember the times when AI got something wrong because it is so accurate. Sometimes AI might have outdated military regulations, or retail store information might not be correct. It's not perfect.

In 1983, Lisanne Bainbridge published a paper in the journal Automatica called Ironies of Automation. Her finding: the more reliable the automated system, the less practiced its human supervisor becomes, and the less able to intervene when the system finally fails. She was studying industrial control rooms.

There is a phrase called "human-in-the-loop." It refers to the importance of supervising AI actions. Keep a human in the loop watching over it. Sometimes humans make mistakes. Sometimes advisors make mistakes.

I've heard from James Pollard that he tested AI by analyzing a financial plan, and it was wrong. I'm not saying that AI isn't a powerful tool. It is simply a powerful algorithm and transformer model, or whatever the technical details under the hood. AI can and does make mistakes. People can get injured by saws and lawnmowers. In the same way, people can get hurt by AI and social media tools. It's a tool. A tool can save a life or take it.

There is still no substitute for human judgment, wisdom, and critical reflection. I would be careful about leaning on it for a decision with significant consequences. At a minimum, you could test the answer with, say, five models, then talk to 2-3 trusted friends or professionals in that field. It depends on, as Nassim Taleb said, how much skin you have in the game.

Medical advice from AI can be risky, tax advice might be risky; however, it does have access to the complex IRS tax code, legal advice can certainly be risky; relationship advice has proven rather helpful - at least for me. Spiritual advice from AI has been excellent. I use a combination of ChatGPT, Manus, and now Ollama to draft devotionals daily. Some might criticize me for that. That's okay. AI devotionals are here to stay. It's not for everybody. AI is not for everybody. Granny uses AI, the small business down the street uses AI. The little leaguer uses AI. The Fortune 500 CEO uses AI. I have the theological training and background. I don't have the time to study bible commentaries for two hours. Sometimes, I must tell the tool to stop using the same Augustine quote. It's not perfect; neither are your devotionals.

AI Is Not for Everybody

Medical advice from AI can be risky, tax advice might be risky, legal advice can certainly be risky; relationship advice has proven rather helpful, at least for me. AI spiritual advice has been excellent. I use a combination of ChatGPT, Manus, and now Ollama to draft devotionals daily. Some might criticize me for that. That's okay. AI devotionals are here to stay. It's not for everybody. AI is not for everybody. Granny uses AI, the small business down the street uses AI. The little leaguer uses AI. The Fortune 500 CEO uses AI. I have the theological training and background. I don't have the time to study bible commentaries for two hours. Sometimes, I must tell the tool to stop using the same Augustine quote. It's not perfect; neither are your devotionals.

Embarrassing but Real

The major reason AI has hurt me is that it trains my brain for quick answers. I've over-relied on AI to build systems instead of prospecting. Perhaps I'm more afraid of the emotional sting of rejection than of wasting time building another AI tool or system. Embarrassing but real. AI has helped me develop and refine strategy for Exponential Advisors, but it can't execute it. Strategy is not a substitute for action.
​

I Would Not Call Them Luddites

I've read several books on AI and written articles about the content that I come across. I've listened to podcasts and read articles by AI "professionals." It's a powerful tool that I think is already revolutionary technology.

Some people are slow, reluctant, or even adamant about not embracing AI. And that's okay. These are usually down-to-earth people who value local community, reading Wendell Berry, and reading books. I wouldn't call them Luddites, although that term gets thrown around all too often. AI is a nascent technology and is rapidly evolving.

I Will Not Hand Over the Keys

I will not hand over the keys to AI. Use it regularly; yes. However, I'm a huge proponent of self-learning. Lifelong learning. Reading books, engaging with people, having intelligent discussions on the street, in the sauna, or at a second-grader's birthday party.

Reading good books is the defense. You don't have to read Dante, Dickinson, and Dostoevsky, or Douglass to develop defensible ideas. Nevertheless, you must read quality material in whatever genre you decide in order to think critically about yourself, others, and the world.

What Happens If It All Gets Shut Down?

A sick child, Claude, and Kit all sit in a room. I take care of the child because he is more important than any post. My youngest son is made in the image of God. I use Claude to review and edit the post I'm about to share about AI. Kit is the email platform that lets me post to all my subscribers with a click of a button. What happens if that all gets shut down? The father and son will still remain on the journey (The Road by Cormac McCarthy).

What do you think about AI? Revolutionary technology, mind-numbing new shiny tool, or somewhere in between?

Do you know a military leader or veteran family thinking through this? Forward it to them.

If someone forwarded this issue to you, you can join Exponential Purpose here: https://exponentialedge.kit.com/profile​

With you in it,

Joshua Brooks, Army Reserve Chaplain

CFP®| Founder, Exponential Advisors

Joshua serves as a chaplain in the U.S. Army Reserve. Nothing here implies endorsement by the Department of Defense, the U.S. Army, or the Chaplain Corps.

Exponential Advisors LLC is an investment adviser registered with the Texas State Securities Board. Registration with any securities authority does not imply a certain level of skill or training.