In other words, there's leverage and value in uniquely, and usefully working with the machine.
If you do things to fit in, you risk being imitated, and potentially replaced. Deviate from the algorithm. There's no better time to be original than now. You'll be rewarded for it.
Imo, two types of people will thrive in the AI era:
1. Generalist Orchestrators - Those who can direct agents better than agents direct themselves.
2. Deep Orchestrators - Those who can think deeper than frontier models.
Everyone else plays catch-up.
Map by @AnthropicAI
Apple keeps raising the ceiling on unified memory. The m5 pro & max is optimised for pro, local inference workflows.
The shift from app to agent experience is coming. Soon, you won't tap through an app to book a bus or pick a movie. Your iOS or Android agent will do it for you.
37signals is such an important voice in the startup world. Proof that you can build sustainable, profitable businesses without going to war.
If you think the ultimate luxury in life is time, their writing and product philosophy is absolute gold.
Enjoy the now. @37signals
Taught an effective and simple framework for GTM motion at @mesaschool_ this evening.
ICP -> Value Proposition -> Distribution -> Pricing -> Positioning
Key takeaway: Positioning is an outcome, not a decision taken in isolation.
Oh, and I love a whiteboard session!
For enterprises deploying AI, knowledge is the true foundation of outcomes. As models grow smarter, organizations must build knowledge thatโs complete, scalable, connected, and self-contained.
In my latest blog, read the 9 principles of AI-ready knowledge.https://t.co/akJHUgp4zV
A new OS and App Store are being built.
@OpenAI announced Apps SDK. Apps embedded within ChatGPT. Interactive and functional through natural language.
The future is conversational. Soon, people will say, "I'm building a ChatGPT app," just like Android or iOS apps.
As build gets more commoditized, distribution and growth will become the new moat.
It's never been easier to ship and iterate. I'm going after hard problems. Watch me.
I've been mentoring undergrads at @mesaschool_ on building AI products. The energy is raw and unfiltered. On campus, I spotted a board where students post sticky notes to find co-founders. To me, these are quiet signals that reveal when a builder era is truly taking shape.
People underestimate taste and distribution in tech. End users don't care about your token optimizer's efficiency or whether your backend is a wrapper - they want their job done well.
You have to tie back engineering to taste. And without distribution, none of this matters.
Context engineering isn't just for AI. It's a life framework.
If you canโt cut the noise and focus on the right context, your decisions wonโt be optimal. You need guiding principles to use the context. And if you don't constantly learn, you don't stay the same - you get worse.
As a consequence of outsourcing writing to AI, @paulg predicted a future with people who can write and people who cannot.
I do not fully understand this future.
But I do know that to write well is to think clearly.
And clear, systemic thought will be an expensive currency.
AI employees (agents) are the future.
With a sloppy 3-liner prompt, I built a functional UI using #GoogleStitch in under 2 minutes.
As execution gets easier, human abilities like emotional intelligence, clarity of thought, and judgement will become the ultimate differentiator.