you searched "how does X work"
read a paragraph that made you search 3 more things
went down a youtube rabbit hole
yak shaving. when you just needed to understand one thing.
type it. get a narrated animation built for exactly that.
https://t.co/fvozNALbgi
you type into Google. results appear in 200 ms across trillions of pages.
it feels magical, but the magic isn't the search. it's what Google already did months ago.
here's how, in 60 seconds:
made with https://t.co/fvozNALbgi
made this using https://t.co/fvozNALbgi
the numbers you cited, at physical scale, with the ₹22,000 Cr / ₹8,000 Cr comparison in the second half.
small correction on the math: one waiver = ~2.75 years at the govt's ₹8,000 Cr figure, not 20. still makes your point.
India mein koi 5 minutes number Google kar le to doodh ka doodh paani ka paani ho jaye.
RBI surplus to Govt : ₹2.87 Lakh Crores (US$30 Bn)
Total listed bank profits : ₹4.11 Lakh Crores (US$42 Bn)
NPCI - which runs UPI pre tax ‘surplus’ : ₹ 1,888 crores (US$200 Mn)
To nuksaan kis ka ho raha hai UPI se aur kaunsi subsidy de rahi hai Govt UPI pe jo chubh rahi hai?
Cost of running ATMs and cash logistics in India is ₹30,500 crores (US$ 3.0 Bn). If you want to optimise shut ATMs and promote UPI instead.
Any levy on UPI is just tax collection. UPI is the one scientific achievement of India everyone acknowledges par ab tax ki bali chadhegi.
agreed on 3D - the fidelity gap is instantly visible to anyone watching.
2D animated explainers are a different ROI curve though. bar is "does the motion teach the concept", not "does it look photoreal". models clear that today.
built https://t.co/uPZjtyv6gf for exactly this. one prompt → explainer in ~2.5 min:
P(all unique) = 365/365 × 364/365 × 363/365 × ...
by person 23, that product has dropped below 0.5 - meaning the chance of a match has crossed 50%.
it doesn't feel like it should. that's the whole point.
23 people in a room.
what's the chance two share a birthday?
your gut says almost zero. the real answer is 50%.
here's what's actually happening on the calendar. watch the collisions:
made with https://t.co/fvozNALbgi
exact same finding for STEM animation. generation is where most tools stop. we made the animation stay editable code, so "redo just scene 2" works without re-prompting the whole video.
both of us are betting on "AI writes structured code" over "AI paints pixels" - feels like the split will define this whole category.
demo of the animation version:
@adamdotnew agentic model + a sandbox with real feedback is the pattern that generalizes. we use the same shape for STEM animation - render the scene, deterministic validators check LaTeX + overflow + overlap, iterate.
different domain, same architecture:
the output, but specifically the rendered artifact. our animator agent writes code, then we render the scene and run deterministic validators (LaTeX errors, canvas overflow, element overlap). animator retries with the validator feedback until it passes.
here's an agent-generated scene that passes:
@andrewchen "solo founders are 100x engineers" - building this from Nepal for the last 8 months. no team, no funding. a text prompt now produces a narrated STEM animation in minutes:
before AI I couldn't have shipped animator + narrator + validator agents as one person.
still opening 3 youtube tabs for one concept?
promptanim turns it into a narrated 1080p animation — lands in under 2 min, the concept actually moving.
2 free credits → https://t.co/uPZjtyv6gf
comment a concept that's never clicked for you. building a few into scenes this week.
the monty hall problem - one of the most unintuitive concepts ever.
here's how promptanim explained it.
every door, every switch, every outcome - in motion.
comment your most confusing concept 👇
type it at https://t.co/zxIN82l40a and watch it come to life.