Where's Charlie Kirk?

How AI just erased him

On each Sunday going forward, my plan is to share with you my adventures in AI.

There are wonderful accounts, like Ryan Stax that will teach you the mindset of building and developing AI for yourself, but for me this is about providing a service.

What I’ve found is that most people don’t actually want to do it themselves.
They want something that’s actionable and also what the technology can do.

For me, that’s valuable.
I get to interact.
I get to see what interests people.

And honestly, my software already works.
And it’s only going to get better.

(Reading headlines to help position your trades is a hidden power).

But here’s the reality for those using ChatGPT for assistance in writing and research…

Most of the AI people use is not in their control, you can’t fully trust it.
You’re better off accepting that now and staying patient every day you use it.

I’ll give you a painful example that happened to me Saturday.

Saturday evening, I was excited to look at my phone and see the response I expected from my insight on finding meaning and covered three meaningful subjects:

Bob Dylan.
Charlie Kirk.
The Sabbath.

And how by learning from each we not only make our life better but our society as well.

Instead, I saw that no one liked my Saturday missive.
And it really hurt.

I poured my soul into that email, and the data backed it up that it stunk.
No reaction.

Later, I saw a new member in a marketing group I joined Profit and Purpose (I suggest anyone reading this pay for whatever tier they can afford) talking about Claude AI.

So I dropped my post into it to get feedback.

Wait.
What?

Charlie Kirk is a major part of my post, and he wasn’t even mentioned.

Friday morning, I had worked on the piece and tried to speed things up with ChatGPT and the bad boy just cut Kirk out.

For some reason, I didn’t do the last edit I normally do.

And apparently, most of the post was erased.

Grok was in on it as well.

Here’s Grok mumbling afterward.

And this is the main reason that when you use AI online — and not something you truly control — you should never assume it will work reliably over time.

Especially since it censors stuff where they think it might by related to politics.

AI drifts.
It loses focus.

I revised the post, and I do believe it has meaning.
Not just personally, but for investing and the future.

The key question I have and maybe you have an answer is that the entire 60s was about protest against the establishment that send young kids to die for what?

Why haven’t we solved this issue?

Having one day set aside to reflect helps everything.

Here’s the revised version:
https://www.ericdealmaker.com/p/the-answer-isnt-in-the-wind

What I enjoy about AI online is not that it can write for me or do research for me.
Those are two things I actually enjoy doing.

(Given the right opportunity, I’ve never been in a situation where I didn’t bring an insight or two that others hadn’t considered).

What excites me is that AI can break down structure.
Writing styles.
Flow.

It gives me an immediate boost — like caffeine — when I get fatigued.

I’ve written emails daily since 2008.
That’s a long time.

It’s tough for anyone in finance to generate research, content, and original insight every day.

AI helps.
But, again, you always have to check its work.

Of course, when AI reviews my work, it likes to give compliments.
Those I do not fact-check.

Here’s what it said when comparing me to one of the most famous newsletter writers, Doug Casey.

Whether that’s half true or not doesn’t matter.

It gives me the confidence and motivation to keep going as I build a loyal readership — one reader at a time.

Hope you’re having a wonderful Sunday morning.

Eric

PS. Here’s how we’re initializing headline testing directly on our own hardware — not just online AI — and training it properly.