Pietro Schirano Profile Banner
Pietro Schirano Profile
Pietro Schirano

@skirano

56,523
Followers
750
Following
1,093
Media
5,409
Statuses

Founder @everartai . 🎨✨ Previously, led AI at @brexHQ . @Uber , @Facebook , and @OpenTable .

San Francisco
Joined September 2009
Don't wanna be here? Send us removal request.
Pinned Tweet
@skirano
Pietro Schirano
5 months
Incredible witnessing the launch of GPT-4o, thanks for having us @sama
Tweet media one
43
51
3K
@skirano
Pietro Schirano
2 years
I don’t care that it’s not AGI, GPT-4 is an incredible and transformative technology. I recreated the game of Pong in under 60 seconds. It was my first try. Things will never be the same. #gpt4
280
1K
27K
@skirano
Pietro Schirano
1 year
This is absolutely wild. I am completely speechless.
Tweet media one
259
1K
11K
@skirano
Pietro Schirano
8 months
Once you know GPT-5 created Sora, everything makes sense.
103
443
9K
@skirano
Pietro Schirano
4 months
GPT-4o is so good at coding that I am having legit mind bending moments back to back. The era of the billion dollar solo entrepreneur really has started.
246
655
9K
@skirano
Pietro Schirano
2 months
$100k monthly revenue in 9 months. First-time founder. Sonnet 3.5 changed everything.
@garrytan
Garry Tan
2 months
Underrated trend
Tweet media one
233
694
8K
86
424
8K
@skirano
Pietro Schirano
3 months
Tweet media one
114
501
6K
@skirano
Pietro Schirano
2 years
Bing AI explains SVB collapse in monkey, banana terms. 🐒🍌
Tweet media one
239
1K
5K
@skirano
Pietro Schirano
1 year
GPT-4 vision truly "hallucinating" on mushrooms. 🍄
Tweet media one
90
516
5K
@skirano
Pietro Schirano
2 years
Using tools like ControlNet, brands can transcend conventional boundaries of expression and explore new dimensions of creativity. Here’s a cool experiment I did featuring the iconic Nike logo. #ControlNet #Nike
52
455
5K
@skirano
Pietro Schirano
1 year
From picture to Lightroom filter settings, thanks to GPT-4 Vision 📸
Tweet media one
Tweet media two
108
346
4K
@skirano
Pietro Schirano
9 months
I am FREAKING OUT. DesignerGPT is listed as number 1 in trending on OpenAI's official announcement for the store! Excuse me as I pass out!
Tweet media one
165
163
3K
@skirano
Pietro Schirano
11 months
Excited to announce my first custom GPT! DesignerGPT 👨‍🎨 Now, you can create beautiful websites directly in ChatGPT, with natively support for dark mode 🌙 Websites hosted remotely via @replit . 🌐 Link in the comments.
128
320
3K
@skirano
Pietro Schirano
2 months
Introducing Claude Engineer 2.0, with agents! 🚀 Biggest update yet with the addition of a code editor and code execution agents, and dynamic editing. When editing files (especially large ones), Engineer will direct a coding agent, and the agent will provide changes in batches.
107
399
3K
@skirano
Pietro Schirano
1 year
Italy welcomes back ChatGPT
Tweet media one
@skirano
Pietro Schirano
1 year
Italy bans ChatGPT
Tweet media one
53
180
2K
54
343
3K
@skirano
Pietro Schirano
1 year
Thanks to GPT-4, I've created a new plugin: StabilityGPT Now, you can run stable-diffusion directly in ChatGPT. 💬🎨 Imagine the possibilities!
126
514
3K
@skirano
Pietro Schirano
5 months
The ChatGPT desktop app just became the best coding assistant on the planet. Simply select the code, and GPT-4o will take care of it. Combine this with audio/video capability, and you get your own engineer teammate.
71
417
3K
@skirano
Pietro Schirano
1 year
I used AI to expand some scenes from Indiana Jones and the Raiders of the Lost Ark. 🎥✨
949
312
3K
@skirano
Pietro Schirano
1 year
"I found him!"
Tweet media one
52
194
3K
@skirano
Pietro Schirano
19 days
Every day getting closer to my dream of building Jarvis thanks to Claude Engineer.
61
199
3K
@skirano
Pietro Schirano
3 months
You're not ready for the new version of Claude Engineer dropping tomorrow.
83
101
3K
@skirano
Pietro Schirano
1 year
DesignerGPT 👨‍🎨 My new plugin that lets you create any website directly in chatGPT! It seamlessly integrates with the stable diffusion plugin, and I didn't code that—GPT just knows how to use both. Absolutely unreal! 🤯 Websites are hosted on a remote server via @replit .
120
372
3K
@skirano
Pietro Schirano
7 months
I used Claude 3 to unredact this part from the OpenAI emails. What's wild is that they used a "per word" redaction, meaning each redaction length is proportional to the length of the words, so assuming context and word length, this is Claude's guess using the page source:
Tweet media one
Tweet media two
121
230
2K
@skirano
Pietro Schirano
1 year
From image to live website using GPT-4 vision and @Replit in less than a minute. Things are about to get so interesting. 🔥
76
328
2K
@skirano
Pietro Schirano
6 months
The biggest alpha in coding with AI right now is: Debug with GPT-4, Code with Claude 3. GPT-4 is still king when it comes to logic, but it's extremely lazy. Meanwhile, Claude would do anything you ask. The duo together is unbeatable.
78
199
2K
@skirano
Pietro Schirano
10 months
Memes come to life with stable video diffusion – mind-blowing! 🔥
45
248
2K
@skirano
Pietro Schirano
1 month
I just updated Claude Engineer to support multiple file/folder creation and edits, simultaneously. It's the closest I've felt to having real-life superpowers. Watch 3.5 Sonnet create a fully functional web app, 6 files, and 5 folders in one shot. Available now in the repo 🔥
74
180
2K
@skirano
Pietro Schirano
2 months
These are AI photos made with FLUX.1 pro "boring snapchat photo circa 2015"
Tweet media one
Tweet media two
Tweet media three
110
129
2K
@skirano
Pietro Schirano
3 months
Introducing Claude Engineer 🧑‍💻 My new repo that allows Sonnet 3.5 to create, read and edit coding files and folders via a simple chat. Think of it as a streamlined local “artifacts” with the added advantage of creating entire projects directly. It even supports online search!
70
202
2K
@skirano
Pietro Schirano
1 year
Italy bans ChatGPT
Tweet media one
53
180
2K
@skirano
Pietro Schirano
25 days
I'm excited to share that I've been partnering with the @AnthropicAI team to build a beautiful, fully customizable customer support AI agent with RAG. Using Claude, NextJS, ShadcnUI, and AWS Knowledge Bases. The entire repo is available on GitHub via the Anthropic Cookbook. ✨
@alexalbert__
Alex Albert
25 days
Announcing the Anthropic Quickstarts repo, a collection of projects designed to help developers quickly get started with building deployable applications using the Anthropic API. Featuring our first quickstart - a Claude-powered customer support agent app:
48
196
2K
57
136
2K
@skirano
Pietro Schirano
4 months
I built my own omni assistant using Gemini 1.5 Flash to guide me through Super Mario 64. Gemini can see what I do on my screen and communicate with me in real time via voice, and thanks to the long 1M context, it has a memory of everything we do together. Incredible.
90
255
2K
@skirano
Pietro Schirano
1 year
Just discovered an amazing GPT-4 trick! 🤯 1. Draw your dream UI in ASCII art. 🎨 2. Let GPT-4 create an accurate 1:1 implementation. ✨ Output in @Replit
Tweet media one
Tweet media two
60
259
2K
@skirano
Pietro Schirano
1 year
Midjourney just launched a new command: /describe. It's basically a super powerful tool for reverse-engineering any image. Use it at your own discretion... 😏
Tweet media one
Tweet media two
37
175
2K
@skirano
Pietro Schirano
2 months
Wow, they really did drop open source Midjourney.
Tweet media one
Tweet media two
Tweet media three
Tweet media four
34
91
2K
@skirano
Pietro Schirano
2 months
After a few days of testing, I can safely say that the Llama 3.1 405B is the best model when it comes to prompting a personality and staying in character. I had a 2 hour long chat with "Richard Feynman" that put a big smile on my face. Here's the prompt I used 👨🏻‍🏫
25
70
2K
@skirano
Pietro Schirano
25 days
I don't think people are ready for the complete revolution of the advertising and model industry that is about to come. This is a model trained on Birkenstock shoes with @everartai and FLUX, using ONLY pictures of the object. The results are instagram ready.
Tweet media one
Tweet media two
Tweet media three
Tweet media four
62
142
2K
@skirano
Pietro Schirano
11 months
My new custom GPT ConvertAnything: The Ultimate File Converter! 🔄 Convert images, audio, videos, PDFs & more with ease. Batch uploads, ZIP support, and easy download links included. A one-stop solution for all conversion needs. Try now! 👇
99
169
2K
@skirano
Pietro Schirano
10 months
Alright this is wild.
25
98
1K
@skirano
Pietro Schirano
10 months
Just incredible. I am running Mistral 8x7B instruct at 27 tokens per second, completely locally thanks to @LMStudioAI . A model that scores better than GPT-3.5, locally. Imagine where we will be 1 year from now. 🏃💨
67
152
1K
@skirano
Pietro Schirano
7 months
Watch Gemini 1.5 Pro collect all my tweets from a video where I scroll through my feed, and neatly organize them into a JSON format. It's even smart enough to discern between tweets, retweets, and quotes. Best scraper tool on the planet! ⚡️
57
162
1K
@skirano
Pietro Schirano
7 months
Claude 3's retrieval ability over long content is so good that if you provide structured data, it essentially acts like a fine-tune! 🤯 I gave it 800 of my tweets and now it just speaks like me. It's hard to convey my level of shock... It even tagged @Replit and @everartai !
Tweet media one
60
135
1K
@skirano
Pietro Schirano
11 months
I am in shock. My model, trained only on Toy Story, generates content from other Pixar films. Type "car", you get Cars. "Monster" and you see Monsters, Inc. "Fish" and Nemo appears. No clue why this happens....
Tweet media one
Tweet media two
Tweet media three
150
56
1K
@skirano
Pietro Schirano
2 months
One of my favorite new hacks has been asking Claude to create a detailed graph of my scripts using Artifacts and then pasting those results directly into the README of the repo.
23
94
1K
@skirano
Pietro Schirano
3 months
I wanted to test the new vision capabilities of Claude 3.5, so I built a full multimodal assistant that can control my Mac. Claude is able to navigate UIs, type, and even send messages, all controlled via voice. This is the future of computing.
53
179
1K
@skirano
Pietro Schirano
6 months
Claude Opus's ability to orchestrate subagents is absolutely insane and deserves more attention. Watch Claude direct subagents to build an entire drawing app. 🧙‍♂️ If you like, I could share this code, which can basically solve any goals you present, step by step. Like Devin!
87
145
1K
@skirano
Pietro Schirano
4 months
"AI won't change the world" @elevenlabsio text to speech:
58
99
1K
@skirano
Pietro Schirano
2 months
This is all the code you need to create a Streamlit app to chat with Llama 3.1 8B locally via @ollama . Fast and private.
25
141
1K
@skirano
Pietro Schirano
2 months
A new open-source image generation model popped out of nowhere and it's actually insanely good?? FLUX.1 by @bfl_ml
Tweet media one
Tweet media two
Tweet media three
Tweet media four
40
127
1K
@skirano
Pietro Schirano
7 months
It happened... Claude 3 generated code GPT-4 could not.
Tweet media one
23
60
1K
@skirano
Pietro Schirano
2 months
Gemini 1.5 Pro Exp is amazing at ingesting and processing large amounts of information. I fed it a screen recording of my viral tweets. In seconds, spit out a complete JSON file with date, the number of likes, shares, views, and comments. 8k tokens output. One shot. No sweat.
25
43
467
@skirano
Pietro Schirano
2 years
I built a Bing AI clone called Zing! Featuring functional GPT-4 and Bing API search integration. I didn't write a single line of code; I just asked, and GPT-4 did the rest 🤯😅 It took roughly 2 hours. Hosted on @Replit
41
152
1K
@skirano
Pietro Schirano
11 months
Steve Jobs is now critiquing my designs directly in Figma! I've just made one of my biggest dreams come true, thanks to GPT-4 Vision + @elevenlabsio . ✨ My Figma window is streamed to GPT, which then provides feedback on the fly. Like on these new design for @everartai
52
146
1K
@skirano
Pietro Schirano
6 months
@infobeautiful @jpohhhh @Leniolabs_ What's interesting is that for humans, 37 and 73 are the most frequently picked numbers.
40
7
1K
@skirano
Pietro Schirano
11 months
Here's another trained model that has blown my mind. 🤯 A Toy Story model that is so good, I can't stop generating stuff. I am confident my algorithm can surpass MidJourney on certain things, especially texture, lighting and styling.
Tweet media one
Tweet media two
Tweet media three
64
87
1K
@skirano
Pietro Schirano
2 months
So many of you have asked, and now it's here 🗣️ Introducing Ollama-Engineer. 🦙 Experience the power of my original agentic coding assistant locally and completely free via @ollama on supported models. Llama 3.1, Mistral Nemo, and more.
28
145
1K
@skirano
Pietro Schirano
1 year
ChatGPT Code Interpreter is incredible! 🌟 Its file uploading & downloading capabilities make it a super flexible tool for various tasks. 💼 Here it extracts colors from an image to create a palette.png 🎨 And automatically compresses large images when running out of memory.
28
81
1K
@skirano
Pietro Schirano
1 year
OMG
Tweet media one
Tweet media two
Tweet media three
@MahmoudEidarous
MahmoudEidarous
1 year
@skirano Can you test it with some complicated questions like that try to provide it with the Whole page and ask him to answer all the questions it will be a bit challenging to the reasoning and extraction of the text from the photo
Tweet media one
1
0
27
44
132
1K
@skirano
Pietro Schirano
6 months
Introducing Maestro ✨ A framework for Claude Opus to orchestrate subagents. Simply ask for a goal, and Opus will break it down and intelligently orchestrate instances of Haiku to execute subtasks, which Opus will review at the end. 🧙‍♂️ The output is saved as markdown 👇
54
155
1K
@skirano
Pietro Schirano
3 months
You asked, I delivered! ⚡️ Introducing Automode for Claude Engineer. When in Automode, Engineer will just go through all the tasks it sets to solve without asking for confirmation. Create multiple files at once, etc. You can enter/exit this mode all within the same chat.
51
96
1K
@skirano
Pietro Schirano
1 year
I fine-tuned an SDXL model on Hiroshi Nagai's art. Just 10 minutes of training and 30 images were enough. The results are wild! 🎨🔥
Tweet media one
Tweet media two
Tweet media three
Tweet media four
67
96
1K
@skirano
Pietro Schirano
2 months
People have been asking me to update Claude Engineer to use gpt-4o-mini, and I thought: Wouldn't it be more fun if I actually showed you how you can use Engineer to create your own custom version that uses gpt-4o-mini as the model? In this video, I go step by step.
30
113
1K
@skirano
Pietro Schirano
7 months
In the era of long-context LLMs, I needed a way to quickly convert a GitHub repo into a text file to feed as a prompt. Introducing RepoToTextForLLMs 📖 Not only does it return the full repo in one file, but it also appends a super prompt for analysis and understanding 👇
81
114
932
@skirano
Pietro Schirano
8 months
Wild! With the new mention capability for GPTs, I can create an app with Grimoire, and seamlessly ask DesignerGPT to deploy it. No change in instructions needed – it's pure intelligence at work. ✨ You can deploy all your ideas for free, with just a mention of @ DesignerGPT.
35
158
968
@skirano
Pietro Schirano
1 year
Never felt as close to the Infinity Gauntlet as I do with this.
Tweet media one
45
42
947
@skirano
Pietro Schirano
3 months
Claude 3.5 Sonnet + Maestro = Sparks of AGI? I asked to make a Mario clone using just geometric shapes, and the wildest part is that it gave the character animations as well, and the shapes seem like novel concepts. It took 3 minutes. Look at the game!
23
109
947
@skirano
Pietro Schirano
1 year
Did @OpenAI just turn on GPT-4 turbo? I'm getting insanely fast replies compared to before.
53
18
932
@skirano
Pietro Schirano
1 year
Using GPT-4 Vision to name never-before-seen architectural styles created with Midjourney. It excels at identifying diverse elements and assigning names to these distinctive creations. 🏛️✨
Tweet media one
27
95
926
@skirano
Pietro Schirano
8 months
This is wild. I gave OS control to GPT-4 via the latest update of Open Interpreter and now it's generating pictures it wants to see in @everartai 🤯 GPT is controlling the mouse and adding text in the fields, I am not doing anything.
69
109
931
@skirano
Pietro Schirano
1 year
I used my StabilityGPT plugin to ask GPT to draw itself, and... 👀 Is this the first self-portrait of ChatGPT? 🪞
Tweet media one
@skirano
Pietro Schirano
1 year
Thanks to GPT-4, I've created a new plugin: StabilityGPT Now, you can run stable-diffusion directly in ChatGPT. 💬🎨 Imagine the possibilities!
126
514
3K
47
135
880
@skirano
Pietro Schirano
2 months
Introducing "Magic Text" 🪄 Change the text in any image while preserving the style and structure of the original one. Only on @everartai 🎨
35
84
887
@skirano
Pietro Schirano
2 years
Why spend time making GPT-4 prompts when it can make the best ones for you? Introducing God mode. Here's how I did it 🧵👇 #GPT4 #AI #GodMode
Tweet media one
27
132
859
@skirano
Pietro Schirano
2 months
5K stars on GitHub for Claude Engineer! ⭐️ Thank you, everyone!
Tweet media one
37
32
861
@skirano
Pietro Schirano
2 months
I've been experimenting with a new AI coding framework that I'll release soon. An evolution of Claude Engineer 🤖 It supports all models, search, and multi-file editing simultaneously. In the video, 3.5 Sonnet (smart and creative) directs Gemini 1.5 Pro (fast and reliable).
44
88
845
@skirano
Pietro Schirano
3 months
A new version of Claude Engineer is out! 🔥 📝 Whole new diff file editing with an improved search and edit function 🌈 Color-coded diffs. 👨‍🏫 The system prompt has been updated with more detailed instructions. 💬 Conversation history management has been improved.
Tweet media one
21
82
822
@skirano
Pietro Schirano
2 months
You are only one prompt away from generational wealth.
56
61
811
@skirano
Pietro Schirano
3 months
Hey Claude Engineer, here's an image. Make a website.
34
81
793
@skirano
Pietro Schirano
1 month
This is such a big deal. Imagine caching entire books worth of content, any coding documentation you want, and retrieving it with multiple API calls. Imagine the sophisticated agents you can now build with this.
@AnthropicAI
Anthropic
1 month
🆕 Prompt caching with Claude. Caching lets you instantly fine-tune model responses with longer and more instructive prompts—all while reducing costs by up to 90%. Available in beta on the Anthropic API today.
172
419
3K
38
48
781
@skirano
Pietro Schirano
4 months
As someone who spends 24 hours a day optimizing the fine-tuning of image models for everyday cases (such as this one), this was hard to watch.
Tweet media one
76
16
775
@skirano
Pietro Schirano
11 months
To test and use GPT-4 Turbo in your apps, just add: 'gpt-4-1106-preview' in the API call!
17
59
756
@skirano
Pietro Schirano
9 months
AI influencers are set to be a major trend in 2024. ⚡️ @everart is the best place to create a consistent AI character. Just a few photos – real or AI-generated – and you can create a unique AI influencer. Let's see how to do it! Meet Emily! 👇
Tweet media one
60
80
756
@skirano
Pietro Schirano
13 days
Here are some o1 prompt techniques I’ve discovered that have worked well for me so far and can likely be applied to any reasoning LLMs. 1. Be explicit about the reasoning chain By requesting a detailed breakdown, you encourage the model to organize thoughts methodically.
Tweet media one
Tweet media two
21
71
780
@skirano
Pietro Schirano
1 year
Thanks to the incredible work of @yoheinakajima , I created BabyAgiGPT 🚀 You can now run autonomous agents directly within ChatGPT. Watch GPT create a 250-page sci-fi novel step by step 📖 @Replit project in the comments
43
130
737
@skirano
Pietro Schirano
2 months
The Midjourney killer, FLUX.1 Pro, is now available to everyone on @everartai .
27
77
758
@skirano
Pietro Schirano
29 days
The number of companies I spoke to, and I mean BIG companies, that are desperate for their engineers to use AI tools, is absurd. Hardest pill to swallow is that engineers, even the good ones, don't do well in this new environment. And it's only going to get worse.
98
55
752
@skirano
Pietro Schirano
3 months
Context caching with Gemini is so good! Here I am caching the entire Gemini Cookbook (around 400k tokens) as an insanely long prompt to create the best Gemini app developer on the planet. Watch Gemini answer any coding questions related to its own APIs.
36
88
745
@skirano
Pietro Schirano
20 days
I added this line in my code editor prompt for Claude Engineer, and now I am seeing a consistent improvement in performance lol
Tweet media one
47
24
724
@skirano
Pietro Schirano
9 months
Balenciaga Pope in Midjourney V6. Unreal.
Tweet media one
34
47
701
@skirano
Pietro Schirano
2 months
Introducing Omni Engineer 🚀 An AI coding framework built for control and effectiveness. Create and edit multiple files in REAL time Web search via DuckDuckGo Save and resume conversations Image support Use any model you want via @OpenRouterAI Here's a quick demo 👨‍💻
42
74
717
@skirano
Pietro Schirano
5 months
I was so impressed with the Astra demo at Google I/O yesterday that I decided to build my own version using Gemini 1.5 Pro Flash. It's so fast and really good. ⚡️ It was even able to detect the gate! Content is streamed directly from my camera. Voice via @elevenlabsio
52
79
700
@skirano
Pietro Schirano
1 year
GPT-4 vision for interior design. 🏠 I love how it's incorporating what it knows about me in the suggestion because of custom instructions. Really incredible technology.
Tweet media one
53
113
689
@skirano
Pietro Schirano
10 months
Just tested @Teknium1 's OpenHermes-2.5-Mistral-7B with a 16k context at q8 – it's exceptional! Minimal loss at 14 tok/s. I've explored many models, but this is the first time I feel like I have the power of ChatGPT right on my local machine – and it's totally free! 🚀
28
65
683
@skirano
Pietro Schirano
4 months
I just got access to Copilot Workspace and it's unreal. I wanted to make a Flask app for Maestro but didn't have the time. As easy as creating a new issue and letting Copilot do the rest. After it writes the code changes, it creates a pull request for you. Great work @github !
25
65
690
@skirano
Pietro Schirano
1 year
Just fine-tuned an SDXL model on a blocky, colorful oil painting style known as 'alla prima'. Used 50 carefully selected images. The results? So beautiful, it's messing with my brain. 🎨
Tweet media one
Tweet media two
Tweet media three
Tweet media four
39
61
679
@skirano
Pietro Schirano
27 days
Here's a tutorial on how you can use Claude Engineer with @Replit . 👨‍🏫 I'll show you how to set up the script with a new or existing project and how you can use it to create any app you want. The magic of using it with Replit is that you can deploy directly from there. ⚡️
21
85
678
@skirano
Pietro Schirano
1 month
The dataset -> The results. We just pushed a major update on @everartai for our FLUX training that will *significantly* increase accuracy and creativity, especially when it comes to training on products. All these pics after the first are fully generated. It's absurd.
Tweet media one
Tweet media two
Tweet media three
Tweet media four
56
46
670
@skirano
Pietro Schirano
1 year
GPT-4 helped me create my dream workflow! 💫 1. Generate a pic with Stable Diffusion 🖼️ 2. Edit using ControlNet to anything I want ✏️ Bonus: It works with Midjourney links too!
27
84
641
@skirano
Pietro Schirano
2 months
This doesn't feel real! When I tell you that @everartai has the best finetuning tech on the planet right now, I am not kidding. Left the dataset, right results being generated in REAL time. We cracked the "brand model" problem, at scale.
28
67
657
@skirano
Pietro Schirano
3 months
Claude Engineer now supports diffs! 🧑‍💻✨ This was one of the most requested features on GitHub. Claude will now know which parts of the code to add or remove as it works on your project, intelligently using diffs. Plus, some other bug fixes.🧹
Tweet media one
25
41
640
@skirano
Pietro Schirano
5 months
im-a-good-gpt2-chatbot it's so good that it created a code interpreter that uses Claude Opus for me. Excuse me as I faint in ontological shock.
27
48
618