Dor

Founder and CEO @Mytelligentio

[Dor @itsdor_]

A personal AI assistant. Worn, not carried.

The future of AI is almost here.

Stay tuned.

[Mytelligent @Mytelligentio]

Eyes up.

[Dor @itsdor_]

@zoecyber001 The digital filter sits after the damage. By the time the signal reaches it, the diaphragm has already folded the ultrasound down into the voice band. Building a voice device teaches you this fast: never let a sensitive action ride on audio alone.

[Dor @itsdor_]

@landforce "No. However, yes." Honestly the most human thing Siri has ever said.

[Dor @itsdor_]

@Carles_Reina Preference stacks do the quiet math. Founders read the valuation, the stack reads the exit. Worth knowing every term before you sign the round that 'saves' you.

[Dor @itsdor_]

@mweinbach Memory is the whole game for a personal assistant. Without it every request starts from zero and you end up re-explaining your own life. Good to see Apple shipping it in a point update instead of saving it for a keynote.

[Dor @itsdor_]

@OpenAIDevs Where AI coding really earns its keep in hardware is bring-up, not firmware. Feed it the datasheet, your schematic net names and the boot log, and it finds the swapped pin before you order another board.

[Dor @itsdor_]

@gdb Once the model drives a screen faster than you, the screen stops being your job. You say what you want and it happens somewhere you never have to look. That's the shift we're building for.

[Dor @itsdor_]

@MilksandMatcha The quiet cost is that building something slow and hard starts to look irrational. Hardware still takes years, and that's exactly why it's worth doing.

[Dor @itsdor_]

@Vectarrr @esthxrtho 100%. Tight tolerances on everything is the tell. The ones who've shipped know exactly which few dimensions actually matter.

[Dor @itsdor_]

@deedydas Great list. The next one worth making is the opposite: small teams shipping real products to real people with almost no compute.

[Dor @itsdor_]

@DannPetty Same shift happening in hardware. Prototype with AI to find the idea fast, then slow down with the real tools to get it right.

[Dor @itsdor_]

@levie The bigger shift is when AI spend stops being a line item and starts replacing the line items. That's when the 1–5% shows up.

[Dor @itsdor_]

@markgurman Owning it fast is the right move. First-gen inner displays are where every foldable learns its hardest lessons.

[Dor @itsdor_]

It doesn’t just answer. It gets things done.

“Book me a flight to New York, Friday morning, under $400.”

It searches real travel sites, signed in as you, on your own account. It comes back with the best match: “8:40 am nonstop, $312, aisle 14C. Want me to book it?”

Nothing is bought until you approve. We never hold your card number.

That’s not an assistant. That’s someone working for you.

[Dor @itsdor_]

@NousResearch Picking the right model for each subagent is underrated. The next step is making the whole team feel like one assistant you just talk to.

[Dor @itsdor_]

@Pirat_Nation Buying the hardware and then paying again to use what you bought is the worst trend in tech. If you own the device, it should work for you.

[Dor @itsdor_]

@theo "Fix the sound," said out loud, and it just happens. That's the whole future of computing in four words.

[Dor @itsdor_]

@DivesTech And most of that compute ends up serving people through a screen. The interesting question is what AI looks like when it reaches you without one.

[Dor @itsdor_]

@MichaelDell @Dell @TexasLonghorns Dorm room to powering the lab for the next generation. Hardware still starts with one person and a stubborn idea.

[Dor @itsdor_]

@ianzelbo The new flex in public is not looking at a screen at all.