Raghav

Building web + mobile apps, one idea at a time | Designing AI content for brands | Geopolitics, policy, gov tech, history | Crypto curious

[Raghav @willtrypromise]

@16vchq We help founders turn rough AI app ideas into mobile screens that look credible enough to ship or pitch, instead of instantly reading as generic AI UI.

[Raghav @willtrypromise]

@TTrimoreau Working on turning rough AI app ideas into mobile screens that feel credible enough to ship, pitch, or test.

[Raghav @willtrypromise]

@anupamrjp Anti-Slop: turn rough AI app ideas into mobile screens people do not immediately dismiss as generic AI UI.

[Raghav @willtrypromise]

@buildinpublic Basically less nice demo and more someone would actually ship this.

[Raghav @willtrypromise]

@sflorimm One narrow painful use case with enough surface area to prove the thing is real. Otherwise you end up building a lot for nobody in particular.

[Raghav @willtrypromise]

@suni_code Be harsh.

[Raghav @willtrypromise]

AI made it cheap to generate screens.

It did not make it cheap to look credible.

You can feel the difference in about 3 seconds:
- generic AI app UI
- same product with actual hierarchy, taste, and trust
- suddenly it looks fundable instead of throwaway

that trust gap is basically what I am obsessed with building around at https://t.co/1KwmWIJwgm

[Raghav @willtrypromise]

AI made it cheap to generate screens.

It did not make it cheap to look credible.

You can feel the difference in about 3 seconds:
- generic AI app UI
- same product with actual hierarchy, taste, and trust
- suddenly it looks fundable instead of throwaway

that trust gap is basically what I am obsessed with building around at https://t.co/1KwmWIK45U

[Raghav @willtrypromise]

@thorstenball A lot of velocity dies in that trust gap.

[Raghav @willtrypromise]

@srotimi_ui Curious what you changed first in that project: hero framing, information hierarchy, or the actual interaction?

[Raghav @willtrypromise]

small reddit update:

the comments that got seen were not product mentions.

they were specific replies to founders asking:
- is this a real signal?
- how do i get referrals?
- how do i make SEO less random?
- where do i find the right creative person?

lesson so far: useful critique beats launch noise.

trying to apply the same thing to Anti-Slop: less look at my tool, more show me where the first screen loses trust.

[Raghav @willtrypromise]

Send me one screenshot of your AI-built app and I will tell you the first thing that makes it look AI-generated.

No generic looks clean feedback.

I mean the actual smell:
- fake depth
- random gradients
- weak hierarchy
- dashboard cards nobody asked for
- copy that sounds like a template

I am doing this because this is exactly the problem I am trying to kill with Anti-Slop.

[Raghav @willtrypromise]

AI made building cheaper.

It did not make first impressions cheaper.

Your first screen still has to explain taste, category, trust, and seriousness before a user reads a single feature.

That is the layer I keep obsessing over with Anti-Slop.

https://t.co/CYErIS1dRr

[Raghav @willtrypromise]

@KieranGilmurray This is right: AI governance has to live inside operating routines, not just policy documents. The question is whether risk reviews, procurement, incident reporting, user feedback, and escalation paths are frequent enough to catch real behaviour after deployment.

[Raghav @willtrypromise]

@alliekmiller this is the part I keep coming back to.

the demo is usually the least important part of the system.

handoffs, failure modes, source of truth, QA, permissions, messy edge cases. boring work is where the product becomes real.

[Raghav @willtrypromise]

@rsalimx the first draft is getting commoditized very fast.

but the uncomfortable part remains: does this feel like the right company, for the right buyer, at the right moment?

that is where design still refuses to die.

[Raghav @willtrypromise]

@markproduct yes, but for different reasons.

AI can make the page. it cannot reliably decide what should be emphasized, what should be removed, or why a buyer should trust it.

I think the job shifts from making pixels to making judgment visible.

[Raghav @willtrypromise]

@fin465 agree, but the bottleneck moved.

building got dramatically cheaper. trust, taste, distribution, and knowing what tiny pain to solve did not.

2026 is amazing for starting, still brutal for being worth keeping.

[Raghav @willtrypromise]

@Varna Food self-reliance debates often turn on the uncomfortable policy mix: procurement incentives, crop choices, import timing, price stability, and farmer risk. The hard part is designing for resilience without creating signals that shift unpredictably every season.

[Raghav @willtrypromise]

The hard part will be making international baselines usable inside domestic institutions. Employment, credit, and security use cases need clear accountability, but also appeal routes, audit capacity, and regulators who can test systems without relying entirely on vendor explanations.