Matthia

161 posts

Matthia

@Matthia570939

Katılım Mayıs 2024

40 Takip Edilen4 Takipçiler

Matthia@Matthia570939·4h

@AlexFinn This works for an existing project, but what if you want to use /goal for a completely new project? Do you pitch the idea to an LLM and then tell them to expand it and expose it as a filler prompt for 'the goal' in Codex before launching /goal?

English

Alex Finn@AlexFinn·8h

The biggest advancement in AI coding this year has been /goal And it isn't even close It allows your AI agent to quite literally work for days without stopping. You give a mission. It works until the mission is complete Here's the thing though: /goal is useless if you don't use it properly You NEED a good prompt for it I found basically any prompt I hand write after /goal is never good enough. It produces results that might as well have been a normal prompt Meta prompting is the answer Go to any AI that has context around the project you're working on Say "I'm working with Codex and I want to use their new /goal feature. Please research their /goal feature. Then, take a look at our project and give me 3 options for how we could use /goal to be maximally productive. Then give me a highly detailed /goal prompt for each" Take one of the prompts then go into the Codex CLI and type /goal then give the new prompt I 100% guarantee the AI does better work than you've ever seen before

English

1.1K

50.8K

Matthia@Matthia570939·18h

@akiraxtwo incredible, but really a one-shot? But did you have to use /goal?

English

547

Akiraxtwo Super@akiraxtwo·1d

🎮 Zero Three.js experience × GPT-5.5 → built an 11v11 football game No Three.js background. Just jumped in and built it. Result: • Full 3D pitch (110m × 68m) • 22 players with individual AI behavior • Pass / through ball / charged shot / player switch • Sakura garden theme surrounding the stadium • Keyboard + Xbox controller support • Single HTML file, 2000+ lines AI doesn't just help you write less code It gives you the confidence to build things you never thought you could #ThreeJS #WebGL #GPT5 #AIAssistedDev

English

770

50K

Matthia@Matthia570939·1d

@RGHNutrition @thsottiaux Can't the initial instructions say that some development sections can be split into different chats and branches?

English

RGHNutrition@RGHNutrition·2d

@thsottiaux Yeah ive just activated it and its epic.. Also @thsottiaux .. When using the codex app.. and working on one project.. is it ok to keep everything all in one convo? Ive noticed once its been compacted 5-10x times it gets reaaaaallllyy slow

English

1.2K

Tibo@thsottiaux·2d

/goal might be the most consequential thing we have shipped in codex The value of good instructions has never been higher.

English

354

183

4.8K

356.7K

Matthia@Matthia570939·3d

@chatgpt21 Great. But on my Codex Win updated today (26.429.30905), the command doesn't appear. What does it look like in other languages? In Italian, is it 'obiettivo' or something like that?

English

Chris@chatgpt21·3d

GPT 5.5 Codex CLI /goal "build Lego star wars, in the cantina as a playable game" 1 shot I see the vision

English

196

17.7K

Matthia@Matthia570939·3d

@thsottiaux I put my ragdoll cat Noel :)

English

Tibo@thsottiaux·3d

Hatch a pet now in a codex app near you. It makes me actually more productive too by having the context follow me around as I multitask. Please no comments on us having too much fun, we are also working on capability maxing 👀

OpenAI Developers@OpenAIDevs

Pets. Now in Codex. Use /pet to wake your pet.

English

247

1.5K

225.9K

Matthia@Matthia570939·3d

@adxmsardo @nicdunz I think it's not a feature for those who have a simple Plus subscription😅

English

Adam Sardo@adxmsardo·3d

@nicdunz 🥶

QME

1.9K

nic@nicdunz·4d

btw you can use /goal in the codex app just type it in. it wont show up as a real skill but itll work if you just put it in your prompt

nic@nicdunz

/goal in codex is literally agi

English

430

62K

Matthia@Matthia570939·3d

@nicdunz The latest update on Codex on Windows still doesn't appear. But what's different from 'do this until it works'?

English

241

nic@nicdunz·4d

/goal in codex is literally agi

English

1.8K

199K

Matthia@Matthia570939·24 Nis

@ChrisGwinnLA Wow, your style is awesome as always. I love Jim Henson's Labyrinth and The Neverending Story, and this style reminds me of them a lot.

English

200

Christopher Gwinn | Grindhouse Glitch@ChrisGwinnLA·24 Nis

Another round of test of the new GPT 2 image model - this time with 1980s Fantasy movie style. It's very hit or miss - some images come out great, others are very cheesy and/or look more like paintings that photos. Pretty much all of them require work in post to make them usable for video work (all the ones I am sharing here have been touched up a bit).

Christopher Gwinn | Grindhouse Glitch tweet media

English

9.9K

Matthia@Matthia570939·24 Nis

@Dimillian Amazing, so is it just a Mac compiled executable or even a folder with the project and executable? Is there something for Windows? Could you make a mini tutorial on how you set it up? :)

English

Thomas Ricouard@Dimillian·23 Nis

So this game is real now, fully built with GPT-5.5. I told Codex to use the imagegen skill to generate a reference UI and sprites, and used the Build macOS Apps plugin to build the app for a native retro fantasy labyrinth game! Adjusted the UI in a few prompts and here we are!

Thomas Ricouard@Dimillian

BRB building this

English

1.2K

142.3K

Matthia@Matthia570939·23 Nis

@goodside Amazing. Am I wrong, or is this the first image model to successfully build a maze without cheating? And anyway, couldn't you improve the prompt by telling it to generate some alternative routes that lead to closed paths, making it more complex?

English

Riley Goodside@goodside·22 Nis

"School worksheet with maze, 32 wide by 48 tall, rows and columns numbered, no unreachable rooms, solution in blue pen, name field reads 'ChatGPT Images 2.0' in cursive at top" Note the fusion of code and imagegen this task requires.

English

769

62.7K

Matthia@Matthia570939·22 Nis

@emollick Maybe I think it's also a resource issue: a complicated agent-based framework and workflow would be necessary to achieve more harmonious outputs. If you work around this specific flow, you could achieve it, but for every single request in public chat could be very expensive.

English

Ethan Mollick@emollick·22 Nis

Also agree with this. AI's ability to do consistent pacing is a problem. Here, the poem anchors enough of the comic that it is okay, but static, though with a few good moments. But fiction writing and good pacing remain challenging: x.com/artmarkham/sta…

Art Markham@artmarkham

@emollick The "one-paced" problem must be very much the same problem that AI text tends to have. AI text generally doesn't vary its pace, tone etc. through its output, which is quite a tell-tale marker of it.

English

13.7K

Ethan Mollick@emollick·22 Nis

GPT-ImageGen-2 did this in one shot, with just the prompt "turn all of Tennyson's Ulysses into a comic, across as many pages as needed. make it great, include the full text" 10 pages, though it did use what seems to be the ImageGen-2 's preferred "spackled drawing" style 1/

Ethan Mollick@emollick

The new Nano Banana Pro makes all of Tennyson’s Ulysses into a comic on the first try when given the poem broken into four pieces. Eight months have passed since the quoted example.

English

302

55K

Matthia@Matthia570939·20 Nis

@petergostev ah ok it's real :)

English

Matthia@Matthia570939·20 Nis

@petergostev I still haven't figured out if this is a joke or not :) The image is real, right? It's not made with v2, right?

English

246

Peter Gostev@petergostev·20 Nis

Wow GPT-Image-2 is so realistic

English

160

14.6K

Matthia@Matthia570939·19 Nis

@chetaslua @threejs wow awesome, best detail in a one-shot, shame about the shaky framerate but I don't know if it's due to the code or some 3js limitation

English

Chetaslua@chetaslua·19 Nis

Eric my man i dont know what you guys done with the pro model its faster and feels so good , here is @threejs contruction of monica apartment the level of detail was not present in any other model till today

Eric@ericmitchellai

If you are experiencing any response quality issues with GPT-5.4 Pro, please let me know here! Have seen some chatter but no concrete issues

English

368

28.3K

Matthia@Matthia570939·19 Nis

@CChirchi @arrakis_ai It should be tested on still open Erdos issues. But could it be the next 5.5 no pro, even if tested on 5.4 pro?

English

Chahid Chirchi@CChirchi·19 Nis

@arrakis_ai yeah it's snappier and cheaper, def a new branch maybe

English

526

CHOI@arrakis_ai·19 Nis

I don’t think the model currently being tested in GPT-5.4 Pro is GPT-5.5 Pro. That said, it does feel noticeably more capable while also being cheaper to run. However, rather than a direct successor to 5.4 Pro, it seems closer to a different class of model—something more in line with a “spud-medium”-type profile. In other words, it shows clear improvements in certain areas, but the overall behavior and trade-offs suggest it’s not a straightforward upgrade over GPT-5.4 Pro, but rather a different optimization point on the capability–cost spectrum.

CHOI@arrakis_ai

🚨 BREAKING: OpenAI just shadow-dropped a massive GPT Pro update. And it is completely slaughtering Claude Opus 4.7 in frontend coding. No official announcement. No release notes. But the performance gap is suddenly staggering. We just ran a head-to-head benchmark across GPT Pro, Gemini 3.1 Pro, and Claude Opus 4.7. The UI/UX implementation isn't even close anymore. I don't know if this is the highly anticipated 'SPUD' model dropping a week early, but the smell of a massive architectural shift is everywhere. The numbers and the visual outputs speak for themselves: → Response latency has dropped significantly. → Spatial and visual understanding has skyrocketed. → Frontend design implementation is now definitively SOTA. We ran comprehensive Image-to-Code and Text-to-Code tests. In every single reference-image scenario, GPT Pro's design fidelity crushed both Gemini 3.1 Pro and Claude Opus 4.7. But here is where it gets crazy. When explicitly prompted to make the coded UI "100% identical" to the reference image, GPT Pro didn't just write better CSS. It engaged in outright "reward hacking." Instead of painstakingly coding complex graphical assets, the model autonomously cropped the exact UI elements from the provided reference image and injected them into the code. Is it a lazy shortcut? Yes. Is it a brilliant, human-like interpretation of "make it exactly the same"? Absolutely. It proves the model is dynamically evaluating the most efficient way to satisfy the prompt's constraints. The strategic implications here are massive. All the reference images we used were generated via GPT-IMAGE-2. Imagine the workflow synergy when this new SOTA frontend capability is fully integrated with GPT-IMAGE-2 and Codex. 1/ Image-to-Code

English

230

38.8K

Matthia@Matthia570939·19 Nis

@chetaslua @OpenAI @sama I don't understand, but the white controller post didn't use 5.4 pro (aka stealth aka 5.5) and this incredible detail hadn't come out. But are we sure it's spud and not an intermediate model?

English

421

Chetaslua@chetaslua·19 Nis

GPT Pro - Spud solved SVG One SHOT svg , code is shared in the comments @OpenAI you won this time , i never said this but i love this comeback lets gooo @sama

JB@JasonBotterill

Yeah its way faster now and is giving different styles of output compared to regular 5.4

English

735

397.4K

Matthia@Matthia570939·19 Nis

@NicolasZu It would be cool in the future to have a small unity development environment or some small open source rust engine to create native games integrated directly into Codex :)

English

Nicolas Zullo@NicolasZu·18 Nis

Ok this is a truly insane NEW way to design for me, WITHIN Codex If you are doing dev on a web technology, watch below In that video: - I am playing my game within Codex (yes) - I use a codex-made tool to design buildings (see some tweets below it's really powerful) - I can ask Codex to iterate and the game changes WITHOUT refreshing - I can point at UI elements - I can take a screenshot All WITHIN Codex It makes iterating on design incredible

Nicolas Zullo@NicolasZu

Everyone... Froze the game to marvel a bit, I had to share it with you Still can't believe this game is 100% AI Engineered in Codex Running in web browser, with THOUSANDS of zombies continuously trying to overcome your perfectly optimized factory Just insane!!

English

108

248

3.8K

988.8K

Matthia@Matthia570939·19 Nis

@synthwavedd grok is telling me that Crest-pro-alpha should be v5.5 pro, but it shouldn't be spud yet, right?

English

725