Getting ChatGPT to generate memes with image and text using memegen.link
Memegen.link is an API that can generate memes based on a requested URL.
The basic format of the URL is https://api.memegen.link/images/<Meme ID>/<The text of the meme here>.jpg
I extracted all possible meme IDs from the meme templates file hosted on memegen. Then, I added them to my prompt, asking ChatGPT to pick the one that best fits the meme it wants to generate, insert it into the URL, followed by the meme caption.
The prompt is basic (included in my other comment for easier visibility) and can probably be further tweaked.
Prompt:
Create a new meme about [Topic]. Use markdown to format the image in this URL: https://api.memegen.link/images/<MEME KEY THAT BEST FITS THE MEME>/<THE TEXT OF THE MEME HERE FORMATTED FOR A URL>.jpg
Meme key - you can use any of them:
Copy the list from this Pastebin and paste it in your prompt: https://pastebin.com/avy3v3Xp
Memegen.link is an API that can generate memes based on a requested URL.
The basic format of the URL is https://api.memegen.link/images/<Meme ID>/<The text of the meme here>.jpg
I extracted all possible meme IDs from the meme templates file hosted on memegen. Then, I added them to my prompt, asking ChatGPT to pick the one that best fits the meme it wants to generate, insert it into the URL, followed by the meme caption.
The prompt is basic (included in my other comment for easier visibility) and can probably be further tweaked.
Prompt:
Create a new meme about [Topic]. Use markdown to format the image in this URL: https://api.memegen.link/images/<MEME KEY THAT BEST FITS THE MEME>/<THE TEXT OF THE MEME HERE FORMATTED FOR A URL>.jpg
Meme key - you can use any of them:
Copy the list from this Pastebin and paste it in your prompt: https://pastebin.com/avy3v3Xp
👍2
Introducing qqbot, a variant of ChatGPT that lives in your IDE.
The cool thing about qqbot is that it knows your codebase.
You can ask it questions like:
- Where is xyz implemented?
- Where are the tests for this function?
- If I want to implement xyz, where do I start?
This is where code is headed. Your IDE should "know" your codebase, and the IDE experience will evolve towards guided, high context interaction with it.
Code isn't going anywhere, but we'll interact with it as more of a personalized entity, not a collection of text files.
This is how we'll make engineers much, much more effective.
Copilot helps you write boilerplate faster, but the most productive engineers are the ones with the most context, who know how to make complex changes, etc.
You don't become a staff eng by typing faster.
qqbot is one person's answer to the prompt: what would it look like to "chat with your codebase"?
Signup form: https://forms.gle/V7Nv7enkQ2L1W6qJ7
The cool thing about qqbot is that it knows your codebase.
You can ask it questions like:
- Where is xyz implemented?
- Where are the tests for this function?
- If I want to implement xyz, where do I start?
This is where code is headed. Your IDE should "know" your codebase, and the IDE experience will evolve towards guided, high context interaction with it.
Code isn't going anywhere, but we'll interact with it as more of a personalized entity, not a collection of text files.
This is how we'll make engineers much, much more effective.
Copilot helps you write boilerplate faster, but the most productive engineers are the ones with the most context, who know how to make complex changes, etc.
You don't become a staff eng by typing faster.
qqbot is one person's answer to the prompt: what would it look like to "chat with your codebase"?
Signup form: https://forms.gle/V7Nv7enkQ2L1W6qJ7
👍2
This media is not supported in your browser
VIEW IN TELEGRAM
Introducing Tensai, your repo-level code assistant
❔ Ask complex questions
✅ Automatically generate PRs for complex tasks
https://TensaiCode.com
The biggest challenges facing developers include:
- Understanding unfamiliar codebases
- Making changes with system-wide impacts
- Making large, repo-wide code edits
Tensai indexes your entire codebase and does the research for you.
Just ask a question in plain english => get an answer => continue the dialogue like in ChatGPT!
Not sure where to start on your next Jira ticket?
Tensai will generate a plan, present it to you, take your feedback, then generate a PR that implements the changes across arbitrarily many files.
This is just the beginning.
Future features include:
- Larger-scale (10k+ files) codebases
- Autogeneration of tests
- AI "linting" - security flaw detection, etc.
- Large-scale code migrations
- Automated feature suggestions
Signup: https://airtable.com/shrhYf24UkDMr2gPl
❔ Ask complex questions
✅ Automatically generate PRs for complex tasks
https://TensaiCode.com
The biggest challenges facing developers include:
- Understanding unfamiliar codebases
- Making changes with system-wide impacts
- Making large, repo-wide code edits
Tensai indexes your entire codebase and does the research for you.
Just ask a question in plain english => get an answer => continue the dialogue like in ChatGPT!
Not sure where to start on your next Jira ticket?
Tensai will generate a plan, present it to you, take your feedback, then generate a PR that implements the changes across arbitrarily many files.
This is just the beginning.
Future features include:
- Larger-scale (10k+ files) codebases
- Autogeneration of tests
- AI "linting" - security flaw detection, etc.
- Large-scale code migrations
- Automated feature suggestions
Signup: https://airtable.com/shrhYf24UkDMr2gPl
This media is not supported in your browser
VIEW IN TELEGRAM
Use ChatGPT to turn anything (text, tables etc.) into HTML code markup
👍6
200 COP in MI: Looking for Circuits in the Wild
Our ultimate goal is to be able to reverse engineer real, frontier language models. And the next step up from toy language models is to look for circuits in the wild. That is, taking a real model that was not designed to be tractable to interpret (albeit much smaller than GPT-3), take some specific capability it has, and try to reverse engineer how it does it. I think there’s a lot of exciting low-hanging fruit here, and that enough groundwork has been laid to be accessible to newcomers to the field! In particular, I think there’s a lot of potential to better uncover the underlying principles of models, and to leverage this to build better and more scalable interpretability techniques. We currently have maybe three published examples of well-understood circuits in language models - I want to have at least 20!
https://www.alignmentforum.org/posts/XNjRwEX9kxbpzWFWd/200-cop-in-mi-looking-for-circuits-in-the-wild
Our ultimate goal is to be able to reverse engineer real, frontier language models. And the next step up from toy language models is to look for circuits in the wild. That is, taking a real model that was not designed to be tractable to interpret (albeit much smaller than GPT-3), take some specific capability it has, and try to reverse engineer how it does it. I think there’s a lot of exciting low-hanging fruit here, and that enough groundwork has been laid to be accessible to newcomers to the field! In particular, I think there’s a lot of potential to better uncover the underlying principles of models, and to leverage this to build better and more scalable interpretability techniques. We currently have maybe three published examples of well-understood circuits in language models - I want to have at least 20!
https://www.alignmentforum.org/posts/XNjRwEX9kxbpzWFWd/200-cop-in-mi-looking-for-circuits-in-the-wild
👍1