LLMs Democratize User Research with Synthetic Audiences

The term 'artificial intelligence' may not fully capture the essence of what we're developing. Consider 'machine intelligence' or even 'alien intelligence' as more fitting descriptors. This emerging intelligence is distinct from our own, though related. It exists in a gray zone where definitions are still being formed. Perhaps renaming it would offer a clearer perspective on its unique nature. YT: https://lnkd.in/grPBwtwN Spotify: https://lnkd.in/gydtsyDU Apple: https://apple.co/3ONwkIB #MachineIntelligence #FutureOfAI #Technology #Innovation #ArtificialIntelligence

[FRANK] Unpopular opinion: Most smartglasses hype has been pure fantasy. 💀 I've been burned so many times. My tech graveyard is full of clunky prototypes & vaporware. I honestly thought the dream of useful smartglasses was dead. But then Xreal, with Google's backing, dropped their latest. And for the first time, I'm genuinely intrigued. 🤔 Their strategy? Not a full PC on your face, but a portable monitor for your phone/PC. That's genius. It's practical, lowers the barrier, and solves a real problem.

Using gpt-realtime-2 to get a glimpse of future voice-first experiences. A market dashboard you don’t click through. You direct it. Say, “Focus on Apple,” and the whole interface changes. Ask, “How did it do over the last 30 days?” and the chart updates. Say, “Go back,” and the market view returns. No menus. No filters. No hunting around. Just intent. What makes this model especially interesting is the interaction loop: you can interrupt it, add more context, change direction, and it keeps reasoning in real time while updating the experience around you. The interface doesn’t ask you to navigate. It just takes you there.

Lately, I’ve been thinking about a metric - the coefficient of humanity. When Cₕ = 1, we are fully present, messy, and alive. When Cₕ = 0, we are merely nodes processing inputs and outputs. Every morning, we face a choice. We can pick up our phones, open our hyper-efficient, algorithmically optimized apps, and reduce our internal friction to zero. Or, we can engage in true human experiences that are inherently high-friction. Streamlining them is to kill their soul. I just published a deep dive on why we need to protect our “uncomputable residue” in a digital-first world. Link is in the comment below👇🏼

Some device details are too small to see clearly. That’s a problem when those details matter. Enlarged models can: Highlight key features Improve understanding in seconds Help start conversations Sometimes seeing it bigger makes all the difference.

New Surface devices are here - and for me, one feature really stands out: the integrated privacy screen. This has been a long time in the making. The goal was clear - deliver strong privacy without compromising the core experience. That means great viewing angles, a bright display, and no added weight or thickness to the device. What makes it even better is the anti-glare treatment. It doesn’t just enhance how the privacy feature performs - it also significantly reduces reflections. And as a bonus, the touchscreen simply feels better to use on an anti-glare surface. Really excited to see this come to life. https://lnkd.in/dhWqSydc

Most CROs ask the wrong pipeline question. "Do we have enough pipeline?" Every Monday morning. Every board meeting. The answer is binary. The actual problem hides one layer below. A team running a 10% disco-to-close win rate has a leak. More pipeline pushes more leads through the same leak. Yannis Sarantis runs revenue ops at Hack The Box (4 million community members). His framework comes from outside revenue ops entirely. He is a former mechanical engineer with a physics background. His instinct is to break complex systems into first principles and look at the layer underneath. The three pipeline questions he runs every Monday at Hack The Box: #1 What percent of ICP leads got first response inside SLA? Cohort by response speed. The slow cohort dies. Fix the speed. #2 What's our week-over-week ICP volume by segment? Total volume up with ICP volume down is a targeting problem. More SDRs will not fix targeting. #3 What's our disco-to-close win rate this month versus last? The most diagnostic metric in the funnel. A drop means competitive shift, pricing, or skill gap. Run them weekly. The leaks surface early. Fixes go in before the quarter is gone. P.S. The full Revenue Vault episode covers Yannis's framework plus the three-layer forecast triangulation he runs at Hack The Box. Watch/listen here: Youtube: https://lnkd.in/gGXSPGCK Spotify: https://lnkd.in/g9bNs3CK Apple: https://lnkd.in/gSTTpPmd

There’s a rhythm to radio that’s hard to explain unless you’ve lived inside it. You learn how to write tight. You learn when a story matters. You learn that local doesn’t mean “small.” It means this is the thing people are talking about in the grocery store, at school pickup, on the drive home. That’s the part I don’t want us to lose. Technology can move fast, and that matters. But radio still needs judgment, tone, and people who understand what belongs on the air. That’s the balance I’m trying to build with Riply. What part of radio do you think technology should never lose?

I came across this social post of mine from 7 years ago, and it still resonated with me and seemed even more relevant with all the AI promises. It never matters how cool or how functionally capable a technology is. What matters is the desire to actually use it to its full potential and apply it to your business problems that matters. You can choose to not transform your business and keep doing what you've always done the way you've always done it, or you can learn a new craft and become a master of it. plus a good Friday pick me up. This was the post: When I'm implementing a new technology or software tool for organizations, I always stress that it isn't the tool that makes it valuable or makes the implementation successful, but the skill with which you use it. I usually use a table saw as an example (ripping 2x4 vs a master cabinet maker), but this works too. True for software, a business system or a 'toy' guitar. Some great blues too! Bonus! https://lnkd.in/gkSq7AyG

How do you get to the point where you just don’t like people? Tech leaders need to answer that one.