some of the smartest people in every field are just faster and better observers than everyone else

The premium on just being someone that stays insanely current on what’s going on in AI is so high right now. The space is changing so fast that you will stand out by being more AI-native and caught up than everyone else. Huge opportunity for the next generation workforce.

The half-life of a v good insight is brutally short. I can remember many of my okay, half-baked insights... but the good stuff disappears unless I write it down.

Voice AI is getting so good over the phone that it’s basically becoming a permissionless API to the enterprise. Enterprises can restrict APIs, but they can’t restrict incoming calls. OpenTable won’t give you API access to make/modify/cancel a reservation? A voice agent can just

Seeing a ton of AI eval/sim startups popping up lately. A year ago, I heard from very few. Now it feels like I get a pitch every day. I see it as a lagging indicator that agentic solutions are finally moving from the lab into real environments. Despite what you might read about

I too am voicepilled. And I think voice is the most underestimated channel right now. My current setup is Wispr Flow and Granola. So much of my day-to-day is becoming voice-first. Over time, this way of working will take hold. It's just a matter of time. And I think

I largely agree with Julian's predictions. Few of us are feeling the model improvements as quickly as we would expect, especially if AI really is doubling the length of tasks it can do every seven months. Devs are probably closest to experiencing these gains. Most others

This is absolutely insane and proof that voice is about to transform the workplace. At Wispr, employees use voice hundreds of times a day to multitask across their entire workflow. Here’s one dev using Cursor and Gmail simultaneously, all through WHISPERING, on a $10 mic.

The daily hype-and-doom AI swings on X, like so many other X things, does not reflect the real world. In companies I talk to, no one is following new developments or thinking of AGI. There is a steady accumulation of valuable use cases & efforts to adjust processes to get more.

The meteoric rise of launch and fundraising videos says a lot about where we are right now. The playing field is flattening. UIs, AI app ideas, and model performance are all converging. More than ever before, an early stage team’s edge is in how you tell the story of what

On Andrej Karpathy's interview with Dwarkesh Patel: I don't understand the current 180. Karpathy was fairly measured. Yes, a single model prompted badly will fail. I see this in my own work all the time. But isolated model performance isn't the market? The actual value is in

Prediction: LLMs in their current form may not be able to do everything, but AI now has enough momentum that this won't matter. Beam engines couldn't do everything either, but they were enough to set off the Industrial Revolution.

I'm hiring an infra-focused FDE at Parloa. We're a voice AI startup revolutionizing customer service, and we're scaling fast. This is a unique role: less customer-facing, more about building repeatable infra for our most strategic clients. Role is foundational for our

Strong hypothesis: therapy will see a huge resurgence with AI. People think there is aversion to talking to AI, but the more human the voice, the more they trust it. And they know there's no judgment involved.

I've done more FDE interviews than I can count. And I've noticed there are four kinds of FDEs in the wild: Connectors (EQ Spike) Builds relationships quickly and drives trust. Born seller and storyteller. Less interest in tackling the hardest technical problems. Strategists

As the application layer converges, in form and function, your only real moat is velocity, where: Velocity (V) = Execution (E) * Ruthless Prioritization (Taste) Teams will only win with a (repeatable) speed advantage.