🚨I’ll be live at 11:30 a.m. ET with Jack🚨
 We’ll share the report with today’s top Income Harvester trades, break down our personal setups, discuss what’s moving markets and more [tap to join us for Market Masters]!
I’ve got to let you in on something that’s completely changed how I approach trading research — and it’s probably not what you’re expecting.
While everyone’s throwing money at ChatGPT subscriptions and fancy AI services, I’m running three separate AI models on a single $20-a-month subscription that work around the clock on everything from trade ideas to my daily Income Harvester report that we share on Market Masters — weekdays at 11 a.m. ET!
These things don’t sleep. They’re literally working 24/7 — researching for my show, tracking trades, coming up with setups and building reports — all while I’m doing other things.
Despite the chatter that AI is some kind of bubble, the real revolution is happening in tangible, daily use — like bringing hedge fund-level analysis to individual traders.
And the best part? It’s all running locally on my machine using open-source models through Ollama. I even share my setup live, screens and all, so others can see exactly what I’m doing without any black-box mystery.
The Model I’m Using and Why It Matters
I’m not just throwing random AI models at the wall. I’ve tested this stuff extensively, and many of these models are remarkably efficient. Some are so lightweight they can run on a cellphone if you give them enough memory. Open-source models give me full control — customizable, offline and able to run anywhere I want, unlike closed, expensive AI platforms.
A lot of people assume these AIs work like chatbots. What I’m working with is totally different. These models treat market prices — candles, open, high, low, close — as their language. They convert that structure into tokens and then predict what the next token should be.
That’s how I’m building my Kronos Price Predictor, which has already shown some impressive early results. Last week, one of its calls nailed the Nasdaq index options close within $2, and it stayed within one standard deviation four out of five days. Of course, this is still experimental. I want to see how it handles real volatility before I rely on it for major trades, but the early signs are remarkable.
Just as companies like Chegg (CHGG) and Duolingo (DUOL) are getting disrupted and destroyed by open AI models, the same dynamic is coming to trading research — except now the power sits in the hands of individuals, not big institutions.
The Hardware Reality Check
Now, full transparency — I’m running this on a Mac Studio with 128GB of RAM, and there’s a reason for that. The Mac architecture allows the entire memory pool to be accessed by the GPU as one unified chunk. When I’m running heavy price prediction workloads, it will suck down every bit of that memory.
I’ve even pushed it to the point where it ran out, but it didn’t crash the machine, which honestly impressed me.
But you don’t need an expensive machine like mine to get started. Lighter versions of these models can run on laptops — or even phones — if you scale them properly. And AMD’s next hardware generation is adopting a similar unified-memory design, which will open the door for Windows and Linux users who want powerful local AI without spending Mac Studio money.
The bigger point is this: The AI infrastructure that once required massive data centers is becoming accessible to anyone willing to learn how to run it. The tools are here, the models are open and if you set things up right, you can build your own research team that works all day and never asks for a raise.
Jeffry Turnmire
Jeffry Turnmire Trading
I host my Morning Monster livestream at 9:15 a.m. ET each weekday on YouTube, and then 30 Minutes of Awesome at 5 p.m. ET each Tuesday!
Please check out my channel and hit that Subscribe button!
You can also follow along and join the conversation for real-time analysis, trade ideas, market insights and more!
- Telegram:https://t.me/+6TdDE7-F6GlhMmJh
Important Note: No one from the ProsperityPub team or Jeffry Turnmire Trading will ever message you directly on Telegram.
I’m just a regular dude in Knoxville, Tennessee: a husband, father, civil engineer, urban farmer, maker and trader.
I’ve been at this trading thing with real money for 20-plus years, and started paper trading over 35 years ago. I have a knack for making some epic predictions that just may very well come true. Why share them? Because I like helping other people — it’s the Eagle Scout in me.
*This is for informational and educational purposes only. There is inherent risk in trading, so trade at your own risk.
P.S. My No. 1 Tech Play Is NOT 1 You’ll Find Making Waves on Earnings…
My No. 1 algo just flagged an incredible tech play for the next week or two…
It’s a stock that’s set to see a wave of inflows from Wall Street as guys read this.
I’ll show you how to get full access to the trade details as well as the engine that flagged the setup today.
