Anshu

@anshuc • 17,390 subscribers

Posting about how to build beautiful apps and products with AI. Let's end slop together. Formerly: YC founder, UI/UX & AI @apple

Shorts

dude GPT-6 Astra is some kind of turbo-AGI machine god for 3D games. It one-shot this in 45 minutes for hardly a couple % of my quota. I figured out how to get great graphics out of it. The trick is image gen. I'll share the process below.

dude GPT-6 Astra is some kind of turbo-AGI machine god for 3D games. It one-shot this in 45 minutes for hardly a couple % of my quota. I figured out how to get great graphics out of it. The trick is image gen. I'll share the process below.

3,050,398 Aufrufe

Jev can draw pictures by predicting every pixel in parallel

Jev can draw pictures by predicting every pixel in parallel

399,626 Aufrufe

I turned my Astra game into a skill! It's called Dream Loop: Astra dreams up the AAA version of your game, then loops till it gets there. Here's a demo I one-shot in 1 hour, 2% weekly quota, no external assets. Prompt and link in replies. Absolutely free. Enjoy!

I turned my Astra game into a skill! It's called Dream Loop: Astra dreams up the AAA version of your game, then loops till it gets there. Here's a demo I one-shot in 1 hour, 2% weekly quota, no external assets. Prompt and link in replies. Absolutely free. Enjoy!

382,145 Aufrufe

I optimized my Dream Loop skill to give you Astra's 3D magic for 80% cheaper! This demo was one-shot, in 25 minutes, for 7% quota on ChatGPT Plus. Yes, the $20/mo one. 3 days of banging my head on the wall and burning Astra tokens. Sharing it for free. Y'all better build some cool stuff with this Link: Details, demo prompt, and cost breakdown are in replies. To be clear: It's not perfect. The demo is a small interactive scene, not a full game. For full games, your best bet is to build it with smaller models, then use this skill to improve the graphics. This is the best I could do, but it's not going to fix the fact that Astra is expensive and Plus quotas are tiny

I optimized my Dream Loop skill to give you Astra's 3D magic for 80% cheaper! This demo was one-shot, in 25 minutes, for 7% quota on ChatGPT Plus. Yes, the $20/mo one. 3 days of banging my head on the wall and burning Astra tokens. Sharing it for free. Y'all better build some cool stuff with this Link: Details, demo prompt, and cost breakdown are in replies. To be clear: It's not perfect. The demo is a small interactive scene, not a full game. For full games, your best bet is to build it with smaller models, then use this skill to improve the graphics. This is the best I could do, but it's not going to fix the fact that Astra is expensive and Plus quotas are tiny

362,360 Aufrufe

Some of you said it still looks like AI slop, and I agreed, so I told Fable to "100x the design." Here's what it did, and all the details on how.

Some of you said it still looks like AI slop, and I agreed, so I told Fable to "100x the design." Here's what it did, and all the details on how.

782,670 Aufrufe

this model saturates the will smith eating spaghetti benchmark so hard, people are making spaghetti eat will smith instead (which is also saturated) source:

this model saturates the will smith eating spaghetti benchmark so hard, people are making spaghetti eat will smith instead (which is also saturated) source:

328,245 Aufrufe

Oh my god Opus 5 is ridiculously good at mobile apps. It might be better than Fable? wtf? I tried my old "conversational calorie tracker" prompts. DAMN this is clean for a v1:

Oh my god Opus 5 is ridiculously good at mobile apps. It might be better than Fable? wtf? I tried my old "conversational calorie tracker" prompts. DAMN this is clean for a v1:

310,431 Aufrufe

This is my "feel the AGI" moment: I used GPT-5.6 Sol to train my own autocorrect model that outperforms GPT-5.6 Sol (wtf??) I have no ML background. I have no idea what I'm doing. I just kept pushing Sol until it spat out a SOTA model. And I spent $0. The motivation: Years of talking to AI have made me terrible at typing. Rather than fix my skill issue, I decided to throw more AI at it. My idea was: instead of autocorrect that interrupts my flow, I want to type fast with mistakes and have AI clean it up after. I wanted the smallest local model possible, for speed, for battery life, for science! So I decided to train my own. Inspired by Andrej Karpathy’s autoresearch, I ran Codex /goal with this setup: pick an experiment, try it, record the results to a doc, throw it out if it fails, and plan the next experiment without repeating failures. I gave a few examples that had to pass, tight latency targets, and let it run. Sol did some amazing things. First, it scanned benchmarks and shortlisted base models: Qwen 3.5, Gemma 4, Liquid LFM 2.5. It found a dataset on HuggingFace for typed text. Then it built a simulator for fingers striking a Mac keyboard, modeling the physical layout with a Gaussian distribution around each key. It simulated striking the wrong key, wrong order, fat-fingering, etc. With the models + data + simulator, it fine-tuned using MLX right on my MacBook. It had a working prototype within an hour! But accuracy was pretty poor. — Problem 1: Tokenization Sol read papers, ran tests, and identified that the tokenizer was the bottleneck. Tokenization makes typos hard for the model to see, so it memorizes mappings instead of using its language priors. Sol tried ByT5, Google’s tokenizer-free byte-level LLM. This made a big improvement, but the model is old and lacked the knowledge needed to reach Sol performance. Sol dug deeper and realized a tokenizer-free model isn’t needed; instead, it used T5Gemma, an encoder-decoder model. This can understand the input deeply before producing output, and furthermore, Sol could post-train the encoder to improve performance. This gave a much higher ceiling. — Problem 2: Loss function Now the model was correcting some typos perfectly, but ignoring most. Sol realized that standard cross-entropy loss was teaching the model to avoid edits, because the vast majority of characters in the training data were left unmodified. The fix was wild: Sol wrote a custom loss function that byte-aligns the source and target strings, uses a dynamic programming algorithm to compute the minimum edits between the two, then weights correct edits much higher than copies. After a lot of tuning, this dramatically improved accuracy. — Problem 3: Autoregression One failure mode remained: if the model made a mistake, it couldn’t backtrack. It could only predict the next token. Teaching it to “think” like a reasoning model would solve this, but would be far too slow. Sol found a beautiful solution: instead of greedily predicting the next token, beam search over all possibilities. This parallelizes the exploration instead of one linear chain-of-thought. At the end, choose the path with highest cumulative log probability. This worked great, but made the experience worse, since the user wouldn’t see progress until the whole search was done. To fix this, Sol made a clever observation: after each search step, the longest common prefix among surviving branches is guaranteed to appear in the final result, so it can be displayed immediately. As the search progresses, weaker paths are dropped and the prefix grows, so the user sees continuous progress. Sol built all this as a custom MLX pipeline that does the parallel decoding on the MacBook GPU, with just ~40ms TTFT. It’s crazy fast and entirely local. — Final eval (error reduction rate, higher is better): - Apple autocorrect: 49.66% - GPT-5.6 Luna: 82.47% - GPT-5.6 Terra: 87.64% - GPT-5.6 Sol: 90.56% - Our model (1.7B): 91.02% Final cost: - 1 quota reset (thanks Tibo) - $0 (And yes, I verified there's no cheating. In fact, we test words scrubbed from the training data to prove the model isn’t memorizing) There were a ton more details and tangents I could write about: contrastive learning, GRPO, DPO, dynamic masking, and more. Sol is a fascinating and creative model. It blew my mind so many times. Don’t let a lack of experience stop you: Sol makes AI experiments accessible to anyone!

This is my "feel the AGI" moment: I used GPT-5.6 Sol to train my own autocorrect model that outperforms GPT-5.6 Sol (wtf??) I have no ML background. I have no idea what I'm doing. I just kept pushing Sol until it spat out a SOTA model. And I spent $0. The motivation: Years of talking to AI have made me terrible at typing. Rather than fix my skill issue, I decided to throw more AI at it. My idea was: instead of autocorrect that interrupts my flow, I want to type fast with mistakes and have AI clean it up after. I wanted the smallest local model possible, for speed, for battery life, for science! So I decided to train my own. Inspired by Andrej Karpathy’s autoresearch, I ran Codex /goal with this setup: pick an experiment, try it, record the results to a doc, throw it out if it fails, and plan the next experiment without repeating failures. I gave a few examples that had to pass, tight latency targets, and let it run. Sol did some amazing things. First, it scanned benchmarks and shortlisted base models: Qwen 3.5, Gemma 4, Liquid LFM 2.5. It found a dataset on HuggingFace for typed text. Then it built a simulator for fingers striking a Mac keyboard, modeling the physical layout with a Gaussian distribution around each key. It simulated striking the wrong key, wrong order, fat-fingering, etc. With the models + data + simulator, it fine-tuned using MLX right on my MacBook. It had a working prototype within an hour! But accuracy was pretty poor. — Problem 1: Tokenization Sol read papers, ran tests, and identified that the tokenizer was the bottleneck. Tokenization makes typos hard for the model to see, so it memorizes mappings instead of using its language priors. Sol tried ByT5, Google’s tokenizer-free byte-level LLM. This made a big improvement, but the model is old and lacked the knowledge needed to reach Sol performance. Sol dug deeper and realized a tokenizer-free model isn’t needed; instead, it used T5Gemma, an encoder-decoder model. This can understand the input deeply before producing output, and furthermore, Sol could post-train the encoder to improve performance. This gave a much higher ceiling. — Problem 2: Loss function Now the model was correcting some typos perfectly, but ignoring most. Sol realized that standard cross-entropy loss was teaching the model to avoid edits, because the vast majority of characters in the training data were left unmodified. The fix was wild: Sol wrote a custom loss function that byte-aligns the source and target strings, uses a dynamic programming algorithm to compute the minimum edits between the two, then weights correct edits much higher than copies. After a lot of tuning, this dramatically improved accuracy. — Problem 3: Autoregression One failure mode remained: if the model made a mistake, it couldn’t backtrack. It could only predict the next token. Teaching it to “think” like a reasoning model would solve this, but would be far too slow. Sol found a beautiful solution: instead of greedily predicting the next token, beam search over all possibilities. This parallelizes the exploration instead of one linear chain-of-thought. At the end, choose the path with highest cumulative log probability. This worked great, but made the experience worse, since the user wouldn’t see progress until the whole search was done. To fix this, Sol made a clever observation: after each search step, the longest common prefix among surviving branches is guaranteed to appear in the final result, so it can be displayed immediately. As the search progresses, weaker paths are dropped and the prefix grows, so the user sees continuous progress. Sol built all this as a custom MLX pipeline that does the parallel decoding on the MacBook GPU, with just ~40ms TTFT. It’s crazy fast and entirely local. — Final eval (error reduction rate, higher is better): - Apple autocorrect: 49.66% - GPT-5.6 Luna: 82.47% - GPT-5.6 Terra: 87.64% - GPT-5.6 Sol: 90.56% - Our model (1.7B): 91.02% Final cost: - 1 quota reset (thanks Tibo) - $0 (And yes, I verified there's no cheating. In fact, we test words scrubbed from the training data to prove the model isn’t memorizing) There were a ton more details and tangents I could write about: contrastive learning, GRPO, DPO, dynamic masking, and more. Sol is a fascinating and creative model. It blew my mind so many times. Don’t let a lack of experience stop you: Sol makes AI experiments accessible to anyone!

180,168 Aufrufe

Whoa, GLM-5.2 is INSANE for UI/UX with the right prompting. Open models have finally closed the gap. You have to push it, but it's super creative. I just asked for a beautiful personal site; this was all the model's idea (multiple turns):

Whoa, GLM-5.2 is INSANE for UI/UX with the right prompting. Open models have finally closed the gap. You have to push it, but it's super creative. I just asked for a beautiful personal site; this was all the model's idea (multiple turns):

178,213 Aufrufe

ok WTF, Fable is absolutely cooking update #3 on my vibe-coded calorie tracker: I didn't like the plain images of food, so Claude turned them into Gaussian splats and rendered them on a virtual dining table???

ok WTF, Fable is absolutely cooking update #3 on my vibe-coded calorie tracker: I didn't like the plain images of food, so Claude turned them into Gaussian splats and rendered them on a virtual dining table???

133,464 Aufrufe

I learned this the hard way: do NOT use SwiftUI if you want your app to look and feel amazing. At least when coding with AI. (sorry, Apple colleagues reading this 😅) I'm sharing my process vibe coding this calorie tracker. I get a lot of questions about the fluid transition in the video. Here's the whole story. Initially, Claude built the grid with SwiftUI. It was quick and easy, and looked good! But the transition to the day view was a boring navigation push/pop. No fun. I wanted something custom. I asked Claude to make it a fluid transition that remaps the food tiles from their source to destination positions. All hell broke loose. Claude tried a bunch of horrible things. Initially it used matched geometry effects, which worked OK but didn't lend themselves well to gesture-driven animations. So it resorted to SwiftUI preference keys + geometry readers to figure out the source and destination positions and calculate the interpolated position based on gesture progress, coordinating across grid and day views. But this meant it had to write a custom layout because it couldn't reposition tiles inside the native SwiftUI grid. And it had to do an awkward handoff between views, which always created ugly pops or jumps. And don't get me started on trying to put it on a bouncy spring, that only made the math 10x buggier. Fortunately, Claude Fable was smart enough to see that this was becoming a disaster (and discover most of the issues itself, in the simulator), so it pivoted away from SwiftUI. Opus might not be so wise, so you'll have to pay attention and intervene. Ultimately, it rewrote it in plain UIKit and everything turned out great. After that, we moved from 2D images to 3D assets, which introduced a new set of performance challenges and yet another rewrite to a single Metal layer, which is what you see below. I can write more about the 2D-to-3D saga if anyone's interested. If I were to do it again, I'd just say "Don't use SwiftUI" from the very first prompt, and save a few hours of headaches. SwiftUI can be amazing for a human iterating directly in code. But agents don't benefit from any of its advantages. Plus, agents have seen decades of UIKit training data, so they're great at writing it, and it's far more flexible. Here's hoping we see more agent-friendly iterations of SwiftUI in the future. Till then, I'm probably going to avoid it.

I learned this the hard way: do NOT use SwiftUI if you want your app to look and feel amazing. At least when coding with AI. (sorry, Apple colleagues reading this 😅) I'm sharing my process vibe coding this calorie tracker. I get a lot of questions about the fluid transition in the video. Here's the whole story. Initially, Claude built the grid with SwiftUI. It was quick and easy, and looked good! But the transition to the day view was a boring navigation push/pop. No fun. I wanted something custom. I asked Claude to make it a fluid transition that remaps the food tiles from their source to destination positions. All hell broke loose. Claude tried a bunch of horrible things. Initially it used matched geometry effects, which worked OK but didn't lend themselves well to gesture-driven animations. So it resorted to SwiftUI preference keys + geometry readers to figure out the source and destination positions and calculate the interpolated position based on gesture progress, coordinating across grid and day views. But this meant it had to write a custom layout because it couldn't reposition tiles inside the native SwiftUI grid. And it had to do an awkward handoff between views, which always created ugly pops or jumps. And don't get me started on trying to put it on a bouncy spring, that only made the math 10x buggier. Fortunately, Claude Fable was smart enough to see that this was becoming a disaster (and discover most of the issues itself, in the simulator), so it pivoted away from SwiftUI. Opus might not be so wise, so you'll have to pay attention and intervene. Ultimately, it rewrote it in plain UIKit and everything turned out great. After that, we moved from 2D images to 3D assets, which introduced a new set of performance challenges and yet another rewrite to a single Metal layer, which is what you see below. I can write more about the 2D-to-3D saga if anyone's interested. If I were to do it again, I'd just say "Don't use SwiftUI" from the very first prompt, and save a few hours of headaches. SwiftUI can be amazing for a human iterating directly in code. But agents don't benefit from any of its advantages. Plus, agents have seen decades of UIKit training data, so they're great at writing it, and it's far more flexible. Here's hoping we see more agent-friendly iterations of SwiftUI in the future. Till then, I'm probably going to avoid it.

118,655 Aufrufe

lmao I always wondered what elon's X notifications look like, so I asked grok 4.5 to build a simulator (link below)

lmao I always wondered what elon's X notifications look like, so I asked grok 4.5 to build a simulator (link below)

75,181 Aufrufe

I don’t think people realize how big a jump Grok 4.6 is. Look at this comparison with Fable on the same prompt. They’re pretty close, but Grok took half the time at ONE TENTH the cost. And I prefer a lot of choices Grok made: - fills the hero image to the top instead of leaving an ugly white strip - flat sharp tiles instead of the usual rounded card slop - fewer pointless labels and elements, more restraint - lighter text, less dense and overwhelming Fable certainly has more depth if you’re trying to do crazy edge-of-distribution stuff. It’s a much bigger model. But for the majority of work, you can get Fable quality for a 90% discount. Just insane how much of a step up it is from Grok 4.5.

I don’t think people realize how big a jump Grok 4.6 is. Look at this comparison with Fable on the same prompt. They’re pretty close, but Grok took half the time at ONE TENTH the cost. And I prefer a lot of choices Grok made: - fills the hero image to the top instead of leaving an ugly white strip - flat sharp tiles instead of the usual rounded card slop - fewer pointless labels and elements, more restraint - lighter text, less dense and overwhelming Fable certainly has more depth if you’re trying to do crazy edge-of-distribution stuff. It’s a much bigger model. But for the majority of work, you can get Fable quality for a 90% discount. Just insane how much of a step up it is from Grok 4.5.

54,718 Aufrufe

Everyone's sleeping on image-to-3D AI models. They can make your app look incredibly unique, with just a little effort. Here's how. This is my calorie tracker, built in a week with nothing but prompting. Just Claude Code + a couple APIs. The visuals are all AI-generated. I'll be sharing the full workflow + all the crazy technical stuff Claude and I did to make this work, so nobody has to struggle through it like me. Deep dive coming soon! Till then, this is the high-level idea: 1. Get a clean image of the food (or whatever your asset is) - In my app, the user describes foods via text, or attaches images (or both) - If text, an LLM extracts the food description and formats it into a specific prompt I tuned for this design, and we generate an image using Z-Image Turbo through fal - If image, we do the same thing but with FLUX.2 [dev] to edit the user image into our reference design - Originally, both used Google Nano Banana, but switching to open models cut costs and latency a ton 2. Gaussian splatting (2D image → 3D model) - I tried various 2D-to-3D options on fal and ended up with TripoSplat as my preferred balance of speed, cost, latency; this turns an image into a 3D model that looks super high quality (link below) - The app displays the 2D image while our backend generates the 3D splat - We "groom" the splat to reduce size and load time by culling low-opacity/scale points 3. Render efficiently on device Originally, it looked great but ran at 10 FPS. Getting to 120 FPS was a crazy journey. TL;DR: - SwiftUI had to go; it forced us to render each asset in independent MTKViews, which wasn't workable - Instead, we composite every dish into one full-bleed CAMetalLayer using MetalSplatter (link below) - We had to make some optimizations within MetalSplatter's code too, to reduce the overhead of sorting points per render Then I added some finishing touches like the subtle rotation and parallax as they move around. I think it turned out pretty cool :) Overall, this took some effort, but we still got it done in less than a day. Hopefully your agent can follow in the footsteps of mine and do it much faster. Keep an eye out for the bigger writeup, which'll give your agent everything it needs. If you have any questions, drop em below!

Everyone's sleeping on image-to-3D AI models. They can make your app look incredibly unique, with just a little effort. Here's how. This is my calorie tracker, built in a week with nothing but prompting. Just Claude Code + a couple APIs. The visuals are all AI-generated. I'll be sharing the full workflow + all the crazy technical stuff Claude and I did to make this work, so nobody has to struggle through it like me. Deep dive coming soon! Till then, this is the high-level idea: 1. Get a clean image of the food (or whatever your asset is) - In my app, the user describes foods via text, or attaches images (or both) - If text, an LLM extracts the food description and formats it into a specific prompt I tuned for this design, and we generate an image using Z-Image Turbo through fal - If image, we do the same thing but with FLUX.2 [dev] to edit the user image into our reference design - Originally, both used Google Nano Banana, but switching to open models cut costs and latency a ton 2. Gaussian splatting (2D image → 3D model) - I tried various 2D-to-3D options on fal and ended up with TripoSplat as my preferred balance of speed, cost, latency; this turns an image into a 3D model that looks super high quality (link below) - The app displays the 2D image while our backend generates the 3D splat - We "groom" the splat to reduce size and load time by culling low-opacity/scale points 3. Render efficiently on device Originally, it looked great but ran at 10 FPS. Getting to 120 FPS was a crazy journey. TL;DR: - SwiftUI had to go; it forced us to render each asset in independent MTKViews, which wasn't workable - Instead, we composite every dish into one full-bleed CAMetalLayer using MetalSplatter (link below) - We had to make some optimizations within MetalSplatter's code too, to reduce the overhead of sorting points per render Then I added some finishing touches like the subtle rotation and parallax as they move around. I think it turned out pretty cool :) Overall, this took some effort, but we still got it done in less than a day. Hopefully your agent can follow in the footsteps of mine and do it much faster. Keep an eye out for the bigger writeup, which'll give your agent everything it needs. If you have any questions, drop em below!

31,119 Aufrufe

Videos

Anya Rossi

sweetdream.ai

SweetDream.ai•Sponsored•Livecam

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Exclusive private shows

1.2k viewers online

Private Show

Join now for exclusive access

Free preview available • Premium content

Finally got Fable 5.5 access, so I made a game based on the worst pun I could think of: POKERMON Pokémon x Balatro. One-shot in 1 hour. Game mechanics, visuals, music, demo video, everything one-shot by Fable. This model is NUTS

Finally got Fable 5.5 access, so I made a game based on the worst pun I could think of: POKERMON Pokémon x Balatro. One-shot in 1 hour. Game mechanics, visuals, music, demo video, everything one-shot by Fable. This model is NUTS

781,805 Aufrufe • vor 24 Tagen

wait actually Claude Code mods are sick. I vibe-coded my own spinner that watches what the main agent is doing and uses Sonnet 5.5 to turn it into little cartoons in real time

wait actually Claude Code mods are sick. I vibe-coded my own spinner that watches what the main agent is doing and uses Sonnet 5.5 to turn it into little cartoons in real time

197,901 Aufrufe • vor 28 Tagen

I ran Claude Opus 5 in one marathon session for 24 hours straight, and ended up with this game. No external assets or code. Every pixel and sound was built by Claude. Prompts, workflow, code, and link below. (sound on!)

I ran Claude Opus 5 in one marathon session for 24 hours straight, and ended up with this game. No external assets or code. Every pixel and sound was built by Claude. Prompts, workflow, code, and link below. (sound on!)

665,231 Aufrufe • vor 3 Monaten

Here's what Astra built in 8 hours. It worked all night on weather effects, Blender modeling, and tons of little details. It spent an hour just adding birds. AGI achieved? Cost so far: ~50% weekly quota. Details below.

Here's what Astra built in 8 hours. It worked all night on weather effects, Blender modeling, and tons of little details. It spent an hour just adding birds. AGI achieved? Cost so far: ~50% weekly quota. Details below.

281,779 Aufrufe • vor 1 Monat

Fable is FINALLY back! Here’s exactly how I used Fable to build this app that got 1M+ views on X. You have everything you need now - no more excuses, go build something incredible!

Fable is FINALLY back! Here’s exactly how I used Fable to build this app that got 1M+ views on X. You have everything you need now - no more excuses, go build something incredible!

578,815 Aufrufe • vor 4 Monaten

Forget threejs demos, Opus 5.5 is *incredible* at building apps. This is Letters Abroad, a language-learning app I'm working on. You exchange letters with an AI pen pal, and learn a language through context (plus some brain science magic). (also built by Opus, of course) I've been tinkering with this the last week. I built a prototype with Astra, but wasn't happy with the design. Tried Fable, still wasn't good enough. Then Opus 5.5 dropped. In a few hours, I've made more progress on the visuals than in a week of burning tokens. This model is insane. It reads my mind and nails it every time. All the gorgeous 3D models and fluid transitions are thanks to Opus. Crazy idea, prompt, repeat. No more quota anxiety, just non-stop flow state. I'll be writing about my prompts/skills/learnings from building this. Lots to share; this generation of models is absolutely game-changing for design!

Forget threejs demos, Opus 5.5 is incredible at building apps. This is Letters Abroad, a language-learning app I'm working on. You exchange letters with an AI pen pal, and learn a language through context (plus some brain science magic). (also built by Opus, of course) I've been tinkering with this the last week. I built a prototype with Astra, but wasn't happy with the design. Tried Fable, still wasn't good enough. Then Opus 5.5 dropped. In a few hours, I've made more progress on the visuals than in a week of burning tokens. This model is insane. It reads my mind and nails it every time. All the gorgeous 3D models and fluid transitions are thanks to Opus. Crazy idea, prompt, repeat. No more quota anxiety, just non-stop flow state. I'll be writing about my prompts/skills/learnings from building this. Lots to share; this generation of models is absolutely game-changing for design!

162,309 Aufrufe • vor 1 Monat

I'm stunned. I told Opus 5.5 to update my app for iPhone Duo, however it wants, and send me a cool demo It ran for 14 hours, redesigned the app to look great in every mode, and made this completely over-the-top video …for 9% quota

I'm stunned. I told Opus 5.5 to update my app for iPhone Duo, however it wants, and send me a cool demo It ran for 14 hours, redesigned the app to look great in every mode, and made this completely over-the-top video …for 9% quota

145,109 Aufrufe • vor 1 Monat

Wow, Claude Fable is absolutely insane for iOS apps. I asked it for an "extremely delightful" calorie tracker. This is from ONE prompt:

Wow, Claude Fable is absolutely insane for iOS apps. I asked it for an "extremely delightful" calorie tracker. This is from ONE prompt:

529,426 Aufrufe • vor 4 Monaten

Fable is pure magic. I wanted a beautiful app to explore ocean wildlife. Fable built this in an hour. It generated videos with Seedance and carefully synchronized them to make these absolutely insane transitions. I've never seen anything like this in an app. Unreal.

Fable is pure magic. I wanted a beautiful app to explore ocean wildlife. Fable built this in an hour. It generated videos with Seedance and carefully synchronized them to make these absolutely insane transitions. I've never seen anything like this in an app. Unreal.

278,983 Aufrufe • vor 3 Monaten

I turned Jev into an image model…? Presenting Jevinci! The paintings are pure JS, no AI image-gen. Jev predicts every pixel's color in parallel and we paint them. The more confident Jev is in a prediction, the wider the brush. (sped up 4x)

I turned Jev into an image model…? Presenting Jevinci! The paintings are pure JS, no AI image-gen. Jev predicts every pixel's color in parallel and we paint them. The more confident Jev is in a prediction, the wider the brush. (sped up 4x)

79,919 Aufrufe • vor 1 Monat

I was curious if Opus 5.5 could come up with a good guitar riff, so I asked it. One thing led to another and it started a band. Introducing… The Context Windows Claude made this music video 100% locally on my MacBook in 24 hours: - 3D modeling and video editing with Blender + Python scripts - images with Qwen-Image 2.1 - music composed from GarageBand instruments on my Mac (Claude reverse engineered the file format) - vocals from an open model called SoulX-Singer, which Claude got running on MLX The way the characters fret and strum the guitars/bass in the video is physically accurate! I don't know what Anthropic is feeding the 5.5 series, but they've made some sort of huge breakthrough in creativity. No other model comes close.

I was curious if Opus 5.5 could come up with a good guitar riff, so I asked it. One thing led to another and it started a band. Introducing… The Context Windows Claude made this music video 100% locally on my MacBook in 24 hours: - 3D modeling and video editing with Blender + Python scripts - images with Qwen-Image 2.1 - music composed from GarageBand instruments on my Mac (Claude reverse engineered the file format) - vocals from an open model called SoulX-Singer, which Claude got running on MLX The way the characters fret and strum the guitars/bass in the video is physically accurate! I don't know what Anthropic is feeding the 5.5 series, but they've made some sort of huge breakthrough in creativity. No other model comes close.

49,652 Aufrufe • vor 27 Tagen

Here's GPT-5.6 Sol vs Fable asked to build the exact same app (prompts in the article!) What do you think? I have my opinions, but I'd like to hear yours. Both built complete, working apps, but with very different approaches: - Sol took less than half the time at the same (high) reasoning - Sol spent much less time on frame analysis. Fable dumped sheets of frames for every transition and scrutinized them multiple times; Sol only checked twice, and both times was just like "looks great, continuing" - At first, Sol intentionally ignored my request to use image generation! It went with procedurally generated food images "for simplicity" (they looked terrible) - On the second prompt, Sol generated images, but intentionally ignored my suggestion for transparent image backgrounds, saying full photos "preserve shadows more convincingly" - Sol's used its built-in GPT Image 2 integration, which saved cost; Fable had to use the Gemini API - Sol didn't seem to know how to use Liquid Glass at all - Fable was more restrained; Sol wanted to add extra labels and detail - On the other hand, Fable decided to add an extra detail page for foods, while Sol did not - Both had nice haptics on device - Fable added a pinch dismiss gesture; Sol did not do gestures

Here's GPT-5.6 Sol vs Fable asked to build the exact same app (prompts in the article!) What do you think? I have my opinions, but I'd like to hear yours. Both built complete, working apps, but with very different approaches: - Sol took less than half the time at the same (high) reasoning - Sol spent much less time on frame analysis. Fable dumped sheets of frames for every transition and scrutinized them multiple times; Sol only checked twice, and both times was just like "looks great, continuing" - At first, Sol intentionally ignored my request to use image generation! It went with procedurally generated food images "for simplicity" (they looked terrible) - On the second prompt, Sol generated images, but intentionally ignored my suggestion for transparent image backgrounds, saying full photos "preserve shadows more convincingly" - Sol's used its built-in GPT Image 2 integration, which saved cost; Fable had to use the Gemini API - Sol didn't seem to know how to use Liquid Glass at all - Fable was more restrained; Sol wanted to add extra labels and detail - On the other hand, Fable decided to add an extra detail page for foods, while Sol did not - Both had nice haptics on device - Fable added a pinch dismiss gesture; Sol did not do gestures

162,141 Aufrufe • vor 3 Monaten

What if your Mac could read your mind? Here’s an early live demo of Marina: personal, proactive intelligence for macOS. There’s no prompting. No chat box. It just understands what you’re working on, and helps you do it faster. Details + beta invites below.

What if your Mac could read your mind? Here’s an early live demo of Marina: personal, proactive intelligence for macOS. There’s no prompting. No chat box. It just understands what you’re working on, and helps you do it faster. Details + beta invites below.

182,410 Aufrufe • vor 4 Monaten

Muse Wonder might beat Astra on benchmarks, but it's nowhere close to Astra on taste. Same prompt: - Wonder (xhigh) = $2.42 in 20 min - Astra (low) = $2.19 in 7 min Benchmarks don't tell you everything 😬

Muse Wonder might beat Astra on benchmarks, but it's nowhere close to Astra on taste. Same prompt: - Wonder (xhigh) = $2.42 in 20 min - Astra (low) = $2.19 in 7 min Benchmarks don't tell you everything 😬

23,253 Aufrufe • vor 17 Tagen

I have an addiction to one-shotting calorie trackers so you know I had to try it on Opus 5.5, and WTF this model is CRAZY One massive prompt plus a tiny followup. Full app, music, voice, editing, all of it. 10% quota on the Max plan. Prompts and full breakdown below.

I have an addiction to one-shotting calorie trackers so you know I had to try it on Opus 5.5, and WTF this model is CRAZY One massive prompt plus a tiny followup. Full app, music, voice, editing, all of it. 10% quota on the Max plan. Prompts and full breakdown below.

33,968 Aufrufe • vor 1 Monat

Fable 5.1 is an absolute beast for game dev. I took my Opus 5 space game and told Fable to "just make it look better". It overhauled everything with a TON of detail, AND made it 2x faster 🤯 Code: Demo: Every pixel you see is made by AI. Fable modeled everything in Blender, generated textures with code, and in a few cases ran image models via CLI. In mere months we'll be vibe-coding AAA graphics at this rate. Just insane. Of course, Fable recorded and edited this comparison video too (with a little help from Opus at the end, because of the Fable outage)

Fable 5.1 is an absolute beast for game dev. I took my Opus 5 space game and told Fable to "just make it look better". It overhauled everything with a TON of detail, AND made it 2x faster 🤯 Code: Demo: Every pixel you see is made by AI. Fable modeled everything in Blender, generated textures with code, and in a few cases ran image models via CLI. In mere months we'll be vibe-coding AAA graphics at this rate. Just insane. Of course, Fable recorded and edited this comparison video too (with a little help from Opus at the end, because of the Fable outage)

54,746 Aufrufe • vor 1 Monat

Wow, if you connect Fable to image-gen and image-to-3D APIs, it's *insane* for game dev. It made an infinite explorable universe, with characters and lore. All assets and sound completely from scratch. 3 prompts in 2 hours, max effort. Workflow below.

Wow, if you connect Fable to image-gen and image-to-3D APIs, it's insane for game dev. It made an infinite explorable universe, with characters and lore. All assets and sound completely from scratch. 3 prompts in 2 hours, max effort. Workflow below.

108,550 Aufrufe • vor 4 Monaten

Well well well, I'm one of the lucky ones with early GPT-5.6 Pro access! One-shot pages from 5.5 Pro vs 5.6 Pro. Can you tell which is which? Looks like OpenAI's finally getting somewhere with design:

Well well well, I'm one of the lucky ones with early GPT-5.6 Pro access! One-shot pages from 5.5 Pro vs 5.6 Pro. Can you tell which is which? Looks like OpenAI's finally getting somewhere with design:

94,488 Aufrufe • vor 4 Monaten

This thing is insane. I ran a web app backwards by turning it into a neural net and doing gradient descent. This is a calculator with handwriting input. It looks like a web app, but it’s actually one neural net running every frame, rendering the app and recognizing handwriting simultaneously. How? I built the basic calculator in JS, compiled it to an NN with Handwired, then spliced a handwriting recognition model directly into the weights. Even more cursed: since the whole app is an NN, we can run it in reverse! Start with the answer, then gradient descent flows backwards through both handwriting and app logic, and finds inputs that produce that answer. WTF Thank you Cheng Lou, this is so much fun!

This thing is insane. I ran a web app backwards by turning it into a neural net and doing gradient descent. This is a calculator with handwriting input. It looks like a web app, but it’s actually one neural net running every frame, rendering the app and recognizing handwriting simultaneously. How? I built the basic calculator in JS, compiled it to an NN with Handwired, then spliced a handwriting recognition model directly into the weights. Even more cursed: since the whole app is an NN, we can run it in reverse! Start with the answer, then gradient descent flows backwards through both handwriting and app logic, and finds inputs that produce that answer. WTF Thank you Cheng Lou, this is so much fun!

11,705 Aufrufe • vor 13 Tagen

Grok 4.5 is INSANE for web dev. I built this scrolling space explorer in just a couple hours with Grok Build. Link below, try it out!

Grok 4.5 is INSANE for web dev. I built this scrolling space explorer in just a couple hours with Grok Build. Link below, try it out!

49,806 Aufrufe • vor 3 Monaten