Welcome to the Friday Wrap-Up for August 14, 2026. This is a short newsletter where I talk about 3 things: What’s on my mind this week, Recommended Reading, and Something Fun. Here's what's on my mind...
On My Mind
Am I wrong about AI-generated YouTube Thumbnails? You likely know how I feel about using AI for creative work (I don't like it). I fear too many business owners, including myself at times, suffer from AI Brian Rot from relying too much on it for certain tasks.
For a while now, one of those tasks was creating YouTube Thumbnails. I worry they don't look real...and some I've seen look nothing like me. I'm so bullish on this that I've told my thumbnail designer to never use AI on them, especially when it comes to images of me or my brand assets.
But maybe that doesn't matter to get the click? I mean, it seems to me creating YouTube Thumbnails is closer to a science than an art. There are countless tools that will evaluate a thumbnail and score it. If all we're trying to do is appease an algorithm, then what's wrong with using a computer program to do it?
But we're not, right? We're trying to convince human beings to think our video is worth watching. And some AI thumbnails are simply god-awful. I mean, look at this one:
The most generous opinion I could muster of it is the image it generated of me a convincing projection of what I might look in 10-15 years, which is similar to my dad. But the dead eyes absolutely kill me.
But again...does it matter? I know it doesn't look exactly like me, but it's close enough. And if it convinces people to click, it's done its job, right?
I'm still unsure. I'm experimenting with them; but while I believe in strong opinions loosely held, I might be holding onto this one a little tighter.
Do you have something on your mind? Let me know at https://streamlinedfeedback.com. You can write in, or leave a voice note. I might even play it on the show. |
Recommended Reading
Superintelligence is a dragon: Over on Platformer, Casey Newton has a great article about Mark Zuckerberg's AI manifesto (shoutout to friend and listener Liz for sending it along).
The truth is anything that references House of the Dragon is good in my book, but the analogy between dragons and AI that Newton makes is salient, and I think very clever. He's also uniquely positioned to make the points he does, as a trustworthy independent journalist and someone close to the tech industry.
I tend to default to, "what's the ulterior motive of this billionaire," and it's nice that Newton states it so plainly.
Something Fun
Schmigadoon!: I finished the far too short series Schmigadoon! this week and thought it was delightful. It's 2 seasons, with 6 episodes each (each season is about the length of a musical), and it's fun and catchy and a testament to great talent. It was also a great palate cleanser after HotD wrapped up.
The worst part about the show is that it didn't get renewed for season 3. If you like musicals, satire, and yelling, "HEY THAT PERSON IS IN THIS!?" this is a perfect show for you. It's available on Apple TV.
Something to Help Your Systems
Call Debriefing: I have a Claude skill called "Daily Call Debrief" that does a few things:
- It runs daily and gathers any calls I had recorded in Google Meet or Fathom.
- Summarizes the calls and adds them to Obsidian
- Adds my action items to Todoist
- Adds Coaching Call and Discovery Call transcripts and adds them to my "Voice of Customer" folder so I can pull from that when I'm working on marketing copy.
AI call recorders is one of the great use-cases for AI in my opinion, because they are recording the call, can crunch data, and do a bunch of stuff with that data. And since you were on the call and present (presumably), you have consumed the primary source and aren't just relying on the summary for the important details.
One thing I like to have it do as well (something I got from my podcast guest Christian) is have it come up with some questions based on the calls. Here's the exact prompt Christian gave listeners:
I want you to pose to me three extremely precise, very hard-hitting questions, all of which are independent and have different coverage, that if sufficiently answered, would yield some kind of transformative, genuinely novel, impactful insight. Please make reference to the primary source material where appropriate in the context of the question.
If you're looking for a solid way to use AI for tasks, this is a great one.
My latest from around the web: