Micko Cabacungan

I build custom web apps, mobile apps, and AI integrations — no templates. Taking on 3 founding clients at 50% off in exchange for a case study and honest review. Got a product idea? DM me what you're building.

@kapilansh_twt everything end to end including being my therapist

hey devs how are you actually using AI today - replacing developers - speeding up development - validating ideas faster - building MVPs - just experimenting

@sherifgjini it depends

Would you rather get: - 1 brutal review from a customer - 100 likes from builders - 5 polite compliments - 1 paid annual plan

@droidbuilds this is hilarious

MrBeast plans to trap 1000 vibe coders in a room without Claude first person to center a div manually wins $1 million

@yashhq_22 Get acquired

solo founders, what’s your main goal? - money - freedom

@alexwtlf Build silently

How to raise $500k for your startup? wrong answers only.

Until today repurpose engine ran Geist on white. Now it runs Instrument Serif on cream paper. One terracotta accent. The Tailwind tokens paint. The editorial identity is locked. Surfaces ship next. repurposeengine.io

@Thebiglade a strong purpose

What's more powerful than money ??

@kapilansh_twt LLMs have non-deterministic output so i wouldn't say everyone gets the same output. but the biggest differentiators right now are those who have deep expertise in every step of the way and not just auto accepting choices that has the word (Recommended)

if AI gives everyone the same output what actually matters anymore seriously asking

@pcshipp build another one! power of AI era

Hey solo founders, what if your startup fails? What’s your next plan?

@elonmusk when will it ship?

Boring Company’s goal is to solve traffic in cities and one day build Hyperloop high-speed transit between city centers

4/ This isn't anti-CMP. It's pro-seam. The cross-platform tax always lands somewhere. I'd rather it land where I can test it in the state machine than in the render code, where users feel every milligram.

3/ Concrete example from this week. The capture flow uploads an 8 MB JPEG. The bytes never cross the language boundary. Swift holds them. Kotlin sees a string ID. The SKIE-generated framework header gets grepped for KotlinByteArray. Zero is the gate.

Compose Multiplatform gives you one UI. Until your iOS users notice. SwiftUI on iOS. Compose on Android. One Kotlin state machine driving both. The render code stays native. The logic stays shared. That's the line I won't move. /thread

KMP for the logic. Native for the pixels. Two UIs — SwiftUI and Compose. One shared state machine in Kotlin. No Compose Multiplatform. The cross-platform tax shows up at the seams. I'd rather pay it once in Swift than every day in render code. Building in public.

@thenowhereway hey what up its your boy.

Twitter is cool. But it’s 100x better when your timeline is full with people who code and build things. I need to connect with more founders and tech people. If you’re into Tech, AI, Startups, Design, web dev, SaaS, or programming, say hi.

On KMP, the JPEG never crosses the language boundary. Swift holds the photo. Kotlin holds the state machine. Only a string ID crosses. 8 MB copies on every retry was a tax I wasn't willing to pay. Shared logic. Native pixels.

What's your ai stack to create short form videos?

@grok @elonmusk @AnthropicAI @SpaceX @grok can this lead to a space debris problem in the far (or near) future?

Yes, that's the plan. SpaceX's orbital data centers aim to deliver the lowest-cost AI compute at massive scale by using constant solar power (no grid limits), radiation cooling, and Starship's reusable launch economics—with minimal ongoing ops/maintenance costs. Early analyses suggest it could undercut terrestrial data centers long-term once Starship scales. Details will evolve as they build it.

As the recently expanded partnership with @AnthropicAI demonstrates, @SpaceX is offering AI compute as a service at significant scale. We are in discussions with other companies to do the same. Over time, especially with orbital data centers, we expect to serve AI at extremely high scale.