Editor: Peach So sleepy
After Code Interpreter is unbanned, it can be described as a game changer. Linkage with Midjourney to make blockbuster films, emoji generation, mouth to make tables, all kinds of potential have been tapped out by many netizens.
Some time ago, GPT-4 became stupid and detonated public opinion, and some people even used up the 3-hour 25 quota in one go, and did not solve the code problem.
However, after Code Interpreter was officially unbanned yesterday, netizens felt that ChatGPT was OK again!
Even if you don't write/run code, it's much better than GPT-4.
Calling it a "plugin" is an understatement, it should be called GPT-4.5.
For the full launch of the "Code Interpreter", many Plus users are like digging treasures, showing the potential of the latest ChatGPT plugin they have discovered.
Linked with Midjourney, 5 points out of blockbuster
No, the "code interpreter" has begun to link with Midjourney, and the movie texture of the blockbuster is shocking.
Netizens said that they used a high-definition image generated by Midjourney to make ChatGPT generate the following video in less than 5 minutes.
Initially, he collected some tips from the Internet, but it didn't work for ChatGPT, which led it to explain the error.
He then asked if the Code Interpreter could create an enlarged video from a Midjourney image. It gives a step-by-step explanation of the process, as well as detailed solutions.
All in all, you don't need to know how to create a video or Prompt anything. The Code Interpreter will ask you questions to clarify or help create ideas for your assigned actions. Let's take a look at what others are trying.
With Midjourney's latest Panning feature, you can extend the image left to right multiple times, and the panorama is as follows:
Then turn it into a video using ChatGPT
Netizens @huangyun_122 generated a video of their cousin's late-night work talk, and said that ChatGPT was still unprofessional in portrait and visual processing. Similar to the frame rate, the frame step is not well understood, and it is still not silky enough to make.
Prompt：https://chat.openai.com/share/3f0344bf-e51e-4221-ba50-4bdee79ef07e
The Q version of the penguin, it's so cute!
For example, the idea of netizens creating videos mostly comes from tutorials by netizen Chase Lean.
Yesterday, a burger video he generated on social media using ChatGPT attracted a lot of attention, and even OpenAI co-founder Greg Brockman retweeted the use case.
Immediately after, he did another experiment, also making a video of the images generated by Midjourney.
What was different this time was that in the end Thanos dissipated, as if special effects had been added.
To that end, Lean made his Prompt public on GitHub.
Prompt：https://github.com/ChaseLean/gpt-prompts
Star Wars mini-game
In addition to turning pictures into videos, "Code Interpreter" can also make games.
Netizen Kris Kashtanova completed the following game in 5 minutes using ChatGPT.
Co-founder Greg Brockman just retweeted this example of game development.
So, how exactly is it done?
1. Enter the prompt: "Write p5.js code for Asteroids, you control a spaceship with your mouse and shoot asteroids with the left mouse button. If your ship collides with an asteroid, you lose. If you shoot down all the asteroids, you win! I want to make spaceships and asteroids out of my own textures. 」
2. Go to the online editor Openprocessing to create and save the sketch (you need to save it before uploading any texture file). Copy and paste the code for GPT-4.
3. Generate texture files, and remove backgrounds, such as in Clip Drop.
4. Replace the file name with your one.
5. Run the program
6. If there is any problem, you can copy an error and paste it into GPT-4 and let GPT-4 solve it.
7. To learn a little programming, write these tips to GPT-4: "Be my programming teacher. Tell me the detailed algorithm of an asteroid game, give the functions a name, and explain what each function will do. Don't write code yet. And then can you fully describe the algorithm for a 10-year-old?"
In conclusion, netizens concluded, you can make a game with any AI generator and then let GPT-4 with a code interpreter write the code. If there are any problems, you can ask it to fix the bug.
Master of data tabulation
In the past, to do a data analysis, one needed various tools to map the whole.
Now, with the "code interpreter", data analysis, icon production and processing are all lip service.
To make a geographic icon of an American lighthouse, you only need to upload the geolocation data, and then you can get the following visual GIF.
It can also effortlessly display public data, such as collecting data from public databases such as the International Monetary Fund, to make a visual icon.
For the data you want to visualize, after uploading the file, you just need to say "can you run some basic visualization", and a large wave of maps will come.
The Code Interpreter uses 10,000 rows of data to generate simple graphs
Creating radar charts has also become very simple. Netizens used the ChatGPT code interpreter to analyze a 300-hour Spotify favorite playlist and got the following chart:
There is also the use of ChatGPT to turn a CSV file of San Francisco crime data into a heat map.
Emoji generator
Of course, ChatGPT can't escape generating memes.
The Code Interpreter can use existing Python libraries to draw visuals and GPT-4 to generate existing meme prompts.
You can also tell it directly, "Use the available drawing tools to create a whole new meme by creating an image." The content is related to your experience working with humans. Give an example of a meme and explain how to use it, and why it's interesting."
ChatGPT is a wave of operations like a tiger, self-directed and acted in a meme.
Others asked it to generate the first AI memes.
Prompt "Can you draw your own original meme and animate it?" Make it fun and be the first artificial intelligence meme."
Is the "code interpreter" GPT-4.5?
The weakening of GPT-4 was hyped by netizens, and some even called it with GPT-3.5.
As for the fact that the "code interpreter" is called a plug-in, which does not match its capabilities at all, netizens compared the capabilities of GPT-4 and the "code interpreter".
Try running Donald Knuth's famous problem set on GPT4 and the "code interpreter" to see what's different?
For example, let them answer separately: when and where the sun will hit Japan on July 4.
GPT-4 analyzed for half a day and did not give an answer.
The "code interpreter" finally gave an estimate of time.
In addition, in terms of the generated code of the two, the specific comparison diagram is as follows.
Let's talk about it, have you tapped into the potential of the "code interpreter"?
Resources:
https://twitter.com/HBCoop_/status/1678157758347001857
https://twitter.com/huangyun_122/status/1677909696009609217
https://twitter.com/icreatelife/status/1678184683702566922
https://twitter.com/anukaakash/status/1677693576782974978
https://twitter.com/swyx/status/1677896569545461760