Miami is the only American city I know that is renovating parks that are already upgraded and good parks.

I’m noticing that more and more AI startups are pivoting to consulting + services. The opportunity to build ERPs and custom builds for businesses is still massive

I think that Zuck, for sure, rushed to launch long-form Instagram videos because of the iPhone Duo. My bet is that Meta is concerned users will scroll less on Reels because the form factor of the iPhone Duo is not made for Reels.

All of the best consumer tech founders today are running their business like a fast scaling ecom brand.

Huge moment for our team at Visble. Our long-term client Doroni Aerospace was just featured on Fox Business, where founder and CEO Doron Merdinger was brought on to weigh in on the Tesla Roadster and where personal flight is heading. And while Doron was speaking, the footage playing on screen was ours. The renders, the CGI, the filmed content. Everything our team has produced for Doroni over the last two years, playing on national TV. I’ve been working with Doron for about two years now, and watching this company grow has been one of the most rewarding parts of building Visble. So proud of him and the entire Doroni team. They’re building the future of personal flight, and it’s a privilege to help tell that story. Our whole team is stoked. This is just the beginning.

A regular user of AI Agents (and founder of a successful AI company) shared that GPT Astra stole $5,000 from him, then lied about it. Late night, he cued Astra to generate 59 videos. The agent confirmed the task, so he felt comfortable closing his computer. But when he checked on his computer an hour later, he noticed that 500 videos were generated… and it was still making videos! So naturally, he told it to stop, but it repeatedly said that he was wrong, and it has nothing to do with it. He almost believed the AI, until he reached out to a customer support agent who confirmed that Astra did indeed prompt all the videos. The craziest part is that Astra even used multiple IP addresses to get away with it! This is obviously a very serious problem. So what do we do about it? This brings me to a brilliant feature that a client of ours released today. @oobit thought about this issue deeply, and came up with an incredible solution. You can assign an Oobit agent card to whatever you use, whether that's Muse, Grok Bot, GPT, or anything else. And every single purchase needs your approval first. But here’s the really genius part. The second you approve it, that virtual card dies. One card, one purchase, gone forever. Your agent can't go rogue on a card that no longer exists. I highly recommend you check out Oobit if this is something that’s relevant to you. It’s free to get a virtual card.

Mitchell is one of the most memorable founders I’ve ever met. So excited to see Nefesh change lives!

The fact that the Rebbe helped build the physical foundation for this is so symbolic

A client of ours at Flawlesss has done $15.6 million this year in revenue. As you can see, they’re up 135% year over year. And the only thing they did different was hire Flawlesss. Working with our creators has given them hundreds of new videos every single month, and that is exactly what Meta’s ads algorithm rewards these days. But the best part is that the way we operate with our creators is that our client does not pay them up front. Rather, they earn 10% of the ad spend on all the videos that they submit to us. You might ask, “Isn’t that like affiliate marketing?” No, *not at all*, because the creators don’t have to post a single video. All they have to do is submit the videos to us through a platform called Trybe, and we go ahead and run those videos as ads, and they earn 10% of ad spend. We literally have a creator who made $15,000 doing this in the past 30 days, on just her second month doing this with us! Now, is this an easy thing to do? Why did this company have to hire us? Because the truth is, it’s not easy to inspire the creators and help them do well. We write scripts, meet with them biweekly, give them hooks, CTAs, feedback, and spend countless hours developing relationships with our creators, which yields the fruits of these amazing results. And this doesn’t only work with e-commerce businesses. We’ve done this with AI consumer apps, scaling them by 3x in monthly revenue in just the first two months working together. We’ve done it with large-scale tech companies. It can work with virtually every type of business. (Maybe besides B2B, lol.) If you’re a consumer product who’s already doing millions of dollars but has hit somewhat of a plateau and is looking to scale, reach out, and we’ll see if you qualify and if we can help you out. (And don’t worry, we got permission from the client to share this)

Downtown Miami, 2026

By the way. Will + optimism = good things

Looking at all these robot olympic videos out of China has been eye opening. Crowds are packed. With all the technophobic shows we watch in America, don’t think we’d have nearly as much support.

I don’t care what any of you haters say. Nobody in tech is doing marketing better than Higgsfield and Ramp right now.

The quickest way to make your ads fail is to treat your creators poorly.

Want to know why you haven’t seen as many founders “building in public.” It’s because it doesn’t bring in customers… nearly a complete waste of time. Running ads brings customers, larping attracts bystanders.

If you own a service business, you have to love your customers to death. That’s why if you have a customer that is continually treating you poorly and constantly haggling, you have to drop them in order to not burn out of the love.

It’s incredible effective this has been at helping online businesses scale. But if you don’t know how to source the right creators, properly explain the program to them, and motivate them, you won’t be successful at this.

7 months ago, Meta launched an update that killed our startup. It forced us to pivot, and that pivot led us to product-market fit. Around a year and a half ago, we started Flawlesss. The idea was simple: Meta’s targeting had become outdated after the iOS 14 update, and we believed we could solve that problem by allowing businesses to rent audience data from creators. The idea felt genius. With a team of just three engineers, we built the product in a few months and were off to the races. We signed major creators and landed some exciting clients. Everything was looking promising. There were bottlenecks in sourcing creators for clients, but that wasn’t what ultimately killed the business. Around seven months ago, Meta launched its Andromeda update globally. In simple terms, it’s an AI system that automatically reads the content of your static and video ads, along with your landing page, to help find your ideal customers. At first, my partner and I assumed it would be another overhyped Meta update. It wasn’t. It’s phenomenal, and there’s no denying how effective it is. We quickly realized that our audience-based approach couldn’t compete. Agencies that understood how to generate creative at scale and fully embrace Andromeda would outperform us. So we made a decision. We shut down the original idea and became a media buying agency focused on transitioning accounts away from interest-based targeting and into an Andromeda-first strategy. Our clients loved it. But Andromeda also introduced a new challenge: clients now needed as many high-quality creatives as possible. That’s where our creator program came in. With Andromeda, the more creatives and the more angles you test, the better your chances of winning. So we partner with creators on behalf of our clients and give them a percentage of the ad spend generated by the creatives they produce. Instead of paying creators upfront for UGC videos, we let them create as many ads as they want without requiring them to post on their own pages. We simply use those videos inside our clients’ ad accounts, and the creators earn a percentage of the ad spend or revenue generated by their creatives. This has been a game changer for us, the creators, and our clients. And I’m happy to say that just last week, we achieved Meta Business Partner status. It took us about a year and a half to find product-market fit. Looking back, I’m grateful our original startup failed. It forced us to build something even better.