Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

small dev experiment today came across Cal AI (recently acquired by MyFitnessPal) 15M downloads, ~$30M/year tried vibecoding it myself on superapp wasn’t gonna pay for the subscription took a few prompts and I had a working v1 lowkey crazy how fast this works

adxtyahq's profile picture

aditya

24,990 subscribers

35,980 görüntüleme • 3 ay önce •via X (Twitter)

OyunEğitimBilim & Teknoloji
LIVE

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

A fan tried to take a Cal players helmet and the equipment guy wasn’t gonna let it happen
0:19

A fan tried to take a Cal players helmet and the equipment guy wasn’t gonna let it happen

Unnecessary Roughness

9,302,214 görüntüleme • 1 yıl önce

Cal AI is a $30M/year company. Now it looks like one.
0:10

Cal AI is a $30M/year company. Now it looks like one.

Studiyo

124,857 görüntüleme • 6 ay önce

a little book i made recently for class, an ode to the beautiful works of ⁦JOSEPH HÄXAN⁩ and an experiment in working circularly
0:15

a little book i made recently for class, an ode to the beautiful works of ⁦JOSEPH HÄXAN⁩ and an experiment in working circularly

dave

18,586 görüntüleme • 2 yıl önce

When I think about myself building my business and how many times: - I was on the brink of failure - I lost a lot of money - I just didn’t think I had what it took I thank myself for never quitting because it made me who I am today.
0:13

When I think about myself building my business and how many times: - I was on the brink of failure - I lost a lot of money - I just didn’t think I had what it took I thank myself for never quitting because it made me who I am today.

JWaller7

29,055 görüntüleme • 7 ay önce

Rawad Akar is working on a self-balancing active ragdoll tool for Unity, and we took a closer look at how it works. ⚙️ See the breakdown: #unity3d
0:17

Rawad Akar is working on a self-balancing active ragdoll tool for Unity, and we took a closer look at how it works. ⚙️ See the breakdown: #unity3d

Jettelly Inc.

28,297 görüntüleme • 3 ay önce

🐶: changbin wrote this a year ago as a birthday present. i recorded it last week by myself and i worked on the tune myself and i mixed it myself and i sent it to hyung. hyung liked it but it's not the final version yet so i'll polish it a bit more and release it for stay~
0:18

🐶: changbin wrote this a year ago as a birthday present. i recorded it last week by myself and i worked on the tune myself and i mixed it myself and i sent it to hyung. hyung liked it but it's not the final version yet so i'll polish it a bit more and release it for stay~

lena ✰

62,178 görüntüleme • 7 ay önce

DeMarcus Lawrence on if a return to the #Cowboys was ever in play: “…I mean everything really happened so fast. The offers came across the table and Dallas offer wasn’t there. I had a decision to make and that’s what happened.” (🎥: Brian Nemhauser on YT)
0:17

DeMarcus Lawrence on if a return to the #Cowboys was ever in play: “…I mean everything really happened so fast. The offers came across the table and Dallas offer wasn’t there. I had a decision to make and that’s what happened.” (🎥: Brian Nemhauser on YT)

Brandon Loree

490,573 görüntüleme • 1 yıl önce

Advertising today couldn’t compete with a jingle like this! Came across this video and it had me 🤣
0:16

Advertising today couldn’t compete with a jingle like this! Came across this video and it had me 🤣

Holly

175,507 görüntüleme • 2 yıl önce

Cal AI might be the most viral health app this year. 8M+ downloads, projected to do $30M revenue this year. Built by two teenagers. Everyone's using it. But nobody's talking about the fact that their AI is completely broken... To the point where users are manually correcting EVERY meal. • Bowl of grapes: 60 cal estimate (actually ~260) • 4 boiled eggs: 1,010 cal estimate (actually ~300) • Meat portions: consistently off by 50% These aren't my words, but the reviews you can see for yourself on App Store and other places. If the whole USP is saving time vs manual logging, and you still have to correct everything... what's the point? What Cal AI did get right: Distribution. • Viral TikTok content • Smart influencer partnerships • 8M downloads in under a year They absolutely crushed GTM. But the product doesn't work lol. The gap between the "90% accuracy" claim vs the actual user experience kills trust. And in health apps, trust is everything. This is the problem with AI apps across the board right now. Everyone's racing to ship fast and go viral. Nobody's asking: "Does this actually solve the problem?" Distribution > product quality is a losing game. As a result, you're bound to encounter problems: 1. Training data doesn't match real-world variety 2. No depth sensing for portion size 3. Poor training data on homemade meals 4. Zero context (is that chicken grilled or fried?) You only get fast & inaccurate answers. This is why, when I started building my own recipe app, I looked at Cal and other AI nutrition apps and noticed that being accurate was the biggest factor. Here's how we're building Nonna differently: ✓ Multi-model AI (different models for different foods) ✓ User feedback loop to improve estimates ✓ Manual override that actually trains the system ✓ Ship when it works, not when it's "good enough" If the AI can't nail it, we're not shipping it. But accuracy alone is boring. So we're also adding some additional features that make you want to use it daily: • Fridge Story: shareable infographic of your fridge contents • Mystery Ingredient: weekly cooking challenges • Cuisine Spin: random inspiration when you're stuck • Expectation vs Reality: before/after photo collages Tl;dr: Distribution gets downloads. Product keeps users. Cal got millions of downloads. How many people still use it daily after manually correcting every meal for a week? Viral marketing with a broken product = expensive way to disappoint people.
0:11

Cal AI might be the most viral health app this year. 8M+ downloads, projected to do $30M revenue this year. Built by two teenagers. Everyone's using it. But nobody's talking about the fact that their AI is completely broken... To the point where users are manually correcting EVERY meal. • Bowl of grapes: 60 cal estimate (actually ~260) • 4 boiled eggs: 1,010 cal estimate (actually ~300) • Meat portions: consistently off by 50% These aren't my words, but the reviews you can see for yourself on App Store and other places. If the whole USP is saving time vs manual logging, and you still have to correct everything... what's the point? What Cal AI did get right: Distribution. • Viral TikTok content • Smart influencer partnerships • 8M downloads in under a year They absolutely crushed GTM. But the product doesn't work lol. The gap between the "90% accuracy" claim vs the actual user experience kills trust. And in health apps, trust is everything. This is the problem with AI apps across the board right now. Everyone's racing to ship fast and go viral. Nobody's asking: "Does this actually solve the problem?" Distribution > product quality is a losing game. As a result, you're bound to encounter problems: 1. Training data doesn't match real-world variety 2. No depth sensing for portion size 3. Poor training data on homemade meals 4. Zero context (is that chicken grilled or fried?) You only get fast & inaccurate answers. This is why, when I started building my own recipe app, I looked at Cal and other AI nutrition apps and noticed that being accurate was the biggest factor. Here's how we're building Nonna differently: ✓ Multi-model AI (different models for different foods) ✓ User feedback loop to improve estimates ✓ Manual override that actually trains the system ✓ Ship when it works, not when it's "good enough" If the AI can't nail it, we're not shipping it. But accuracy alone is boring. So we're also adding some additional features that make you want to use it daily: • Fridge Story: shareable infographic of your fridge contents • Mystery Ingredient: weekly cooking challenges • Cuisine Spin: random inspiration when you're stuck • Expectation vs Reality: before/after photo collages Tl;dr: Distribution gets downloads. Product keeps users. Cal got millions of downloads. How many people still use it daily after manually correcting every meal for a week? Viral marketing with a broken product = expensive way to disappoint people.

Denislav Jeliazkov

37,989 görüntüleme • 8 ay önce

Custom Grass Renderer Recently, I had a chance to create a custom grass renderer for Dinolords Here`s how it works #indiegames #strategygame
0:18

Custom Grass Renderer Recently, I had a chance to create a custom grass renderer for Dinolords Here`s how it works #indiegames #strategygame

Kacper Szwajka

44,184 görüntüleme • 1 yıl önce

i saw a tweet on this and was lowkey feeling myself after i did it mama is a speedrunner now get my ass on the leaderboards and gdq
0:18

i saw a tweet on this and was lowkey feeling myself after i did it mama is a speedrunner now get my ass on the leaderboards and gdq

ain 🍉 @ currently playing: ocarina of time

32,247 görüntüleme • 17 gün önce

This is a small game prototype that I've been working on off and on for the last few weeks. Basically it should be a kind of Gainer Lifesim, because I want to build something like that for years...
0:18

This is a small game prototype that I've been working on off and on for the last few weeks. Basically it should be a kind of Gainer Lifesim, because I want to build something like that for years...

inkboyjosh

48,049 görüntüleme • 2 yıl önce

Annotations in Dev Mode is now live! As a Figma designer, I’ve always had a love-hate relationship with marking up a file. Today, I’m writing about how this feature came to be — for both design and dev. Scroll 🧵 and/or read the blog:
0:10

Annotations in Dev Mode is now live! As a Figma designer, I’ve always had a love-hate relationship with marking up a file. Today, I’m writing about how this feature came to be — for both design and dev. Scroll 🧵 and/or read the blog:

Oscar Nilsson

46,504 görüntüleme • 2 yıl önce

right now my reinforcement learning model is basically doing this and i came up with a solution for it all by myself (by stealing it from pufferlib) that not only solves it, but also helps keep training stable and fast
0:10

right now my reinforcement learning model is basically doing this and i came up with a solution for it all by myself (by stealing it from pufferlib) that not only solves it, but also helps keep training stable and fast

kache

60,094 görüntüleme • 9 ay önce

My uber picked me up in a Tesla and fell asleep on the highway. I’ve driven with FSD and trust how well it works so I wasn’t worried. It took the exit and drove me to my destination all by itself. He was asleep for 90% of the 20 minute drive haha
0:19

My uber picked me up in a Tesla and fell asleep on the highway. I’ve driven with FSD and trust how well it works so I wasn’t worried. It took the exit and drove me to my destination all by itself. He was asleep for 90% of the 20 minute drive haha

Kyle Wade

1,290,125 görüntüleme • 1 ay önce

I haven't had a Brutalist Room Update for y'all in a few days because I've been working on this behemoth: More pics/video once I finish UVing and detailing it.
0:10

I haven't had a Brutalist Room Update for y'all in a few days because I've been working on this behemoth: More pics/video once I finish UVing and detailing it.

Anton Hand

14,386 görüntüleme • 3 yıl önce

One year ago today, I had zero video editing or AI experience. This post took off and ignited a creative passion I never expected. So… here’s a fun remake 😆
0:18

One year ago today, I had zero video editing or AI experience. This post took off and ignited a creative passion I never expected. So… here’s a fun remake 😆

𝕏erias

12,665 görüntüleme • 6 ay önce