@Jason West yeah, the repeat issue also happens because many outlets pick up the same story and do their spin on it, too. My GPT has a lot of my history, and business building blocks known, so the results I get are kind of tailored to me, as it knows my vibe. So, each day my brief opens with a theme, then I get the 5 different topics, then a "what matters most" section at the end, wrapping up the report, theme and takeaways. The headlines are interesting to me, as I almost always roll my eyes at them.... meaning, how the news gets spun my different orgs drives me nuts. So, for me, it's prob the angles for content that I could then plug into Claude to help flesh out more content. I will typically paste the entire brief, with sourced links, too... and my Claude is really dialed in on my voice, and ICP. For fun, here is one of todays 5 briefs. AI is getting increasingly good at interpreting CGM data. That creates a new problem: false precision. A 2026 research project called GlucoFM trained an artificial-intelligence foundation model on more than 109,000 hours of continuous glucose monitor data. Researchers designed the model to separate longer-term physiological patterns from short-term glucose events and reported improved prediction of diabetes risk, insulin resistance, and beta-cell dysfunction across several datasets. This is a preprint, not yet something to use clinically. But the direction is worth watching. Plain English Today a CGM mostly tells you: “Here's what your glucose did.” Tomorrow, AI may increasingly say: “Here's what this pattern probably means.” And eventually: “Here's what may happen next.” That sounds incredibly useful. It could be. Contrarian angle More sophisticated interpretation creates the illusion that glucose contains the entire metabolic story. It doesn't. An AI can find patterns in glucose that a human would never notice. But it still does not automatically know: - what you ate accurately; - your muscle mass; - why you slept poorly; - whether you're stressed; - whether you're ill; - what your medications changed; - whether the sensor is wrong; - what outcome actually matters to you.

@Julie Helmer absolutely!

@Jason West Yeah... I gave a very basic prompt for a line of copy I would be sharing online, and I received two different images from GPT. One came up with a pretty cool concept. Then I uploaded a couple of reference photos of me and asked it to create a movie poster... and it went pretty wild. The detail in the copy is pretty awesome.

@John Santos gotta keep it real. 😂 That was a good tip, though.

I am looking forward to seeing how this works once I have access. Hopefully get it soon. Prior to this, I have used GPT quite a bit to review my health data, so I am curious to see if things are improved or not.

Looking forward to trying it out. I have been using Gemini/Banana for a while now... it's hard to go back to GPT. But this does look promising.

@Steffen Braun, I have looked in to Abacus previoulsy... may be worth checking out again one of these days.

I downloaded it, but apparently it will not run on my older MacBook Pro. Perhaps the Macs must be new and AI-enabled?