正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Merlin is the first text editor that can actually edit your text 🧙‍♂️ Think Grammarly, but supercharged with LLMs 🔋 It uses GPT-4 to: - understand your writing - ask clarifying questions - propose improvements - use those improvements to magically fix your essay Here's a demo:

Varun Shenoy

7,005 subscribers

412,082 次观看 • 3 年前 •via X (Twitter)

科学技术教育艺术

Anya Rossi• Live Now

Private livecam show

11 条评论

Varun Shenoy 的头像

Varun Shenoy3 年前

Try it out for yourself: An OpenAI API key is all you need. Merlin explores the intersection of LLMs and writing in a way that preserves the author's voice and creativity. This is just a prototype of what is possible. Let me know what you think!

Corinne Marie Riley 的头像

Corinne Marie Riley3 年前

WOOO

pavan sethi 的头像

pavan sethi3 年前

yess

Danny Raede 的头像

Danny Raede3 年前

Now do books

David Lam 的头像

David Lam3 年前

Congrats! Amazing!

Elisson Amboni 的头像

Elisson Amboni3 年前

English only?

Varun Shenoy 的头像

Varun Shenoy3 年前

Multilingual support is fairly strong with GPT-4 and pretty good with GPT-3!

Commander Zimba🎖🇬🇭 的头像

Commander Zimba🎖🇬🇭3 年前

@readwise save this

Simon Assion 的头像

Simon Assion3 年前

@DeepLcom Write is doing the same.

Shubham Patil 的头像

Shubham Patil3 年前

What is the accuracy rate here (guessing it's pretty good because of gpt-4)

Varun Shenoy 的头像

Varun Shenoy3 年前

Hard to judge what accuracy would be in this context, but GPT-4 gives much richer suggestions than GPT-3.

相关视频

PhD Students – Use this tool to paraphrase the text in your manuscript for FREE. First, why you need to paraphrase? The first draft you write is often very rough. It needs to be revised to polish it for submission. For this, you can use a paraphraser. Here is one you can use it for FREE. 1. Go to and log in. 2. Upload your file or copy/paste your text in a new file. 3. From the options on right, click Paraphraser. 4. Now select the text you want to paraphrase. 5. After text selection, click paraphrase. 6. Trinka.ai will show the paraphrased text. 7. You can also see the changes made. 8. You can increase or decrease the level of paraphrasing. 9. You can also select to show the revisions. 10. Go through the paraphrase text carefully. 11. You can choose among multiple paraphrased alternatives. 12. After your final selection, click replace. 13. This will replace original text with paraphrased text. 14. You can apply this process for the entire text. Make sure that you go through the changes made. This is important so that only your text gets refined. However, the meaning of the text remains the same. In addition to paraphrasing, Trinka also offers to ➝ Fix all grammatical errors in your manuscript ➝ Proofread your manuscript before submission ➝ Check consistency in your manuscript writing ➝ Evaluates the level of bias in your writing ➝ Checks plagiarism level in your manuscript Try Trinka.ai here: Anything you’d like to add?

PhD Students – Use this tool to paraphrase the text in your manuscript for FREE. First, why you need to paraphrase? The first draft you write is often very rough. It needs to be revised to polish it for submission. For this, you can use a paraphraser. Here is one you can use it for FREE. 1. Go to and log in. 2. Upload your file or copy/paste your text in a new file. 3. From the options on right, click Paraphraser. 4. Now select the text you want to paraphrase. 5. After text selection, click paraphrase. 6. Trinka.ai will show the paraphrased text. 7. You can also see the changes made. 8. You can increase or decrease the level of paraphrasing. 9. You can also select to show the revisions. 10. Go through the paraphrase text carefully. 11. You can choose among multiple paraphrased alternatives. 12. After your final selection, click replace. 13. This will replace original text with paraphrased text. 14. You can apply this process for the entire text. Make sure that you go through the changes made. This is important so that only your text gets refined. However, the meaning of the text remains the same. In addition to paraphrasing, Trinka also offers to ➝ Fix all grammatical errors in your manuscript ➝ Proofread your manuscript before submission ➝ Check consistency in your manuscript writing ➝ Evaluates the level of bias in your writing ➝ Checks plagiarism level in your manuscript Try Trinka.ai here: Anything you’d like to add?

Faheem Ullah

17,951 次观看 • 1 年前

Mermaid improvements incoming! - Proper code editor when you want to get your hands dirty and go manual - Prominent errors, with a potential auto-fix (don't expect much, yet) - Improved Mermaid renderer (fixed text wrapping, and better styling support)

Mermaid improvements incoming! - Proper code editor when you want to get your hands dirty and go manual - Prominent errors, with a potential auto-fix (don't expect much, yet) - Improved Mermaid renderer (fixed text wrapping, and better styling support)

Excalidraw

23,306 次观看 • 3 个月前

You can now double click on text with the refinement tool to edit text live in your preview.

You can now double click on text with the refinement tool to edit text live in your preview.

v0

76,642 次观看 • 1 年前

Don't use ChatGPT to just summarise text - It can do so much more! This mindblowing prompt will turn your text into a DIAGRAM! Here's a tutorial how to do it for FREE: (It even works inside your Obsidian notes) 👇

Don't use ChatGPT to just summarise text - It can do so much more! This mindblowing prompt will turn your text into a DIAGRAM! Here's a tutorial how to do it for FREE: (It even works inside your Obsidian notes) 👇

Ilya Shabanov

192,045 次观看 • 3 年前

📣 New feature alert! 📣 Meet Pages, your new editing BFF! No need to start from scratch - just hit "Edit this Response," highlight text, and tap "Ask" to tweak, expand, or polish your writing. Use it now at

📣 New feature alert! 📣 Meet Pages, your new editing BFF! No need to start from scratch - just hit "Edit this Response," highlight text, and tap "Ask" to tweak, expand, or polish your writing. Use it now at

Microsoft Copilot

38,170 次观看 • 1 年前

Google Nano Banana annotations has arrived to Freepik. Edit your images with text boxes to specific precise edits to your images. You can then bring this into AI video for flawless shots. Here's how easy it is to do it!

Google Nano Banana annotations has arrived to Freepik. Edit your images with text boxes to specific precise edits to your images. You can then bring this into AI video for flawless shots. Here's how easy it is to do it!

Jerrod Lew

41,745 次观看 • 9 个月前

Don't think your vote matters? Your vote matters A LOT. Use it to save our country. Canada is broken but we can fix it!

Don't think your vote matters? Your vote matters A LOT. Use it to save our country. Canada is broken but we can fix it!

Right Blend

20,548 次观看 • 1 年前

A demo of the animations[.]dev skill file using Claude Code. It can review your animations, suggest improvements, and basically answer all your motion-related questions based on my experience gathered over the years.

A demo of the animations[.]dev skill file using Claude Code. It can review your animations, suggest improvements, and basically answer all your motion-related questions based on my experience gathered over the years.

Emil Kowalski

123,952 次观看 • 5 个月前

We think text-to-image AI is pretty interesting, so here's text-to-BIM! It won’t “create a museum in the style of Zaha Hadid.” (Yet.) But you can describe your building and get an editable 3D model in return. Coming soon from Hypar !

We think text-to-image AI is pretty interesting, so here's text-to-BIM! It won’t “create a museum in the style of Zaha Hadid.” (Yet.) But you can describe your building and get an editable 3D model in return. Coming soon from Hypar !

Hypar

28,706 次观看 • 3 年前

If you’re a politician or in the public eye, it is really important to add ALT text to images in your tweets especially those with text. ALT text lets blind, visually impaired and disabled users access your content via screen readers.

If you’re a politician or in the public eye, it is really important to add ALT text to images in your tweets especially those with text. ALT text lets blind, visually impaired and disabled users access your content via screen readers.

Rob McDowall AMRSPH FRSA

21,246 次观看 • 1 年前

hey! we've built a new AI playground over at @cloudflare to demonstrate what can be achieved by chaining multi-modal models together. think — audio → text → image → text, or composing multiple LLMs together. here's a quick demo video!

hey! we've built a new AI playground over at @cloudflare to demonstrate what can be achieved by chaining multi-modal models together. think — audio → text → image → text, or composing multiple LLMs together. here's a quick demo video!

harley turan

70,711 次观看 • 1 年前

This is the inside of a GPT/LLM as it reads your text and "thinks" about what to say next

This is the inside of a GPT/LLM as it reads your text and "thinks" about what to say next

Brandon

43,353 次观看 • 1 年前

I DID IT 🚀 First text to video demo is live. For the first time, the visuals actually match the script. Still rough, lots to fix, but honestly… I think I’m on the right path. What do you think? PS: The prompt used to generate the video is in the comments 🙌

I DID IT 🚀 First text to video demo is live. For the first time, the visuals actually match the script. Still rough, lots to fix, but honestly… I think I’m on the right path. What do you think? PS: The prompt used to generate the video is in the comments 🙌

Mihai  Youtube Automation

38,595 次观看 • 5 个月前

Andrej Karpathy builds a personal wiki to think with LLMs. I took that idea and went further. Claude Code + Obsidian = an AI that actually knows you. Your goals. Your context. Your history. Not a chatbot. A second brain that remembers. Here's exactly how to build it:

Andrej Karpathy builds a personal wiki to think with LLMs. I took that idea and went further. Claude Code + Obsidian = an AI that actually knows you. Your goals. Your context. Your history. Not a chatbot. A second brain that remembers. Here's exactly how to build it:

leopardracer

216,495 次观看 • 2 个月前

Embedding features learned with sparse autoencoders can make semantic edits to text ✨ (+ a reading/highlighting demo) I've built an interface to explore and visualize GPT-4 labelled features learned from a text embedding model's latent space. Here's a little video, more in 👇

Embedding features learned with sparse autoencoders can make semantic edits to text ✨ (+ a reading/highlighting demo) I've built an interface to explore and visualize GPT-4 labelled features learned from a text embedding model's latent space. Here's a little video, more in 👇

Linus

51,614 次观看 • 2 年前

A lot of talk about ai 🤖 … I hope this clip helps a lot of you … if you want to ask your questions on “Tea with GaryVee” on text me at 1-212-931-5731 ask the question on text and use hashtag #askgaryvee at the end of the text … #ai #artificialintelligence #garyvee

A lot of talk about ai 🤖 … I hope this clip helps a lot of you … if you want to ask your questions on “Tea with GaryVee” on text me at 1-212-931-5731 ask the question on text and use hashtag #askgaryvee at the end of the text … #ai #artificialintelligence #garyvee

Gary Vaynerchuk

22,323 次观看 • 1 年前

Essay is focused on providing a structured environment that helps you develop the skills to become a better communicator. Essay has unique tools that encourage productive writing habits, and make it easy to plan, test, and play with your writing. Try Essay for free:

Essay is focused on providing a structured environment that helps you develop the skills to become a better communicator. Essay has unique tools that encourage productive writing habits, and make it easy to plan, test, and play with your writing. Try Essay for free:

Essay

102,795 次观看 • 1 年前

Editing your videos in @GooglePhotos is easier than ever. You can now: 1. Create shareable videos quickly with templates 2. Try the easy-to-use redesigned video editor 3. Add the perfect soundtrack 4. Tell your story with custom text 5. Customize individual video clips

Editing your videos in @GooglePhotos is easier than ever. You can now: 1. Create shareable videos quickly with templates 2. Try the easy-to-use redesigned video editor 3. Add the perfect soundtrack 4. Tell your story with custom text 5. Customize individual video clips

Google

63,426 次观看 • 6 个月前

Typing just became… Typeless. Meet Typeless 1.0.2 for Mac — a tool that transforms your voice into clear, accurate writing across your Mac. Speak naturally and let Typeless handle the typing, corrections, and structure for you. With Typeless, you can dictate, translate, or ask for quick edits in any language or accent. It converts your speech into polished text up to 10× faster than traditional typing, while automatically fixing mistakes along the way. --- 1️⃣ Dictation Typeless acts as a powerful voice keyboard that works across all applications on your Mac. When you speak, it understands your intent, organizes your ideas, and converts your natural speech into well-structured writing. Whether you're drafting emails, notes, documents, or messages, Typeless helps you turn spoken thoughts into clean text instantly. Controls - Press Fn to start or stop dictation - Hold Fn for quick, short dictation --- 2️⃣ Translation Typeless makes writing in other languages effortless. You can speak in your native language and have Typeless translate your words instantly into the language you want. This allows you to communicate, write, and respond in foreign languages smoothly and naturally. Controls - Press Fn + Space to start translation - Press Fn to stop translation --- 3️⃣ Ask Anything Typeless Typeless also lets you interact with your text using voice commands. You can select any text and simply say how you want it changed. Typeless can edit, rewrite, answer questions, or perform quick actions based on your request, making editing and improving text much faster. Controls - Press Fn + Space to start Ask Anything - Press Fn to stop Ask Anything --- With Typeless, your voice becomes the fastest and easiest way to write, edit, and communicate on your Mac. Your voice is now your keyboard. Get Typeless → Available now on Mac, Windows, iOS, and Android. #Typeless

Typing just became… Typeless. Meet Typeless 1.0.2 for Mac — a tool that transforms your voice into clear, accurate writing across your Mac. Speak naturally and let Typeless handle the typing, corrections, and structure for you. With Typeless, you can dictate, translate, or ask for quick edits in any language or accent. It converts your speech into polished text up to 10× faster than traditional typing, while automatically fixing mistakes along the way. --- 1️⃣ Dictation Typeless acts as a powerful voice keyboard that works across all applications on your Mac. When you speak, it understands your intent, organizes your ideas, and converts your natural speech into well-structured writing. Whether you're drafting emails, notes, documents, or messages, Typeless helps you turn spoken thoughts into clean text instantly. Controls - Press Fn to start or stop dictation - Hold Fn for quick, short dictation --- 2️⃣ Translation Typeless makes writing in other languages effortless. You can speak in your native language and have Typeless translate your words instantly into the language you want. This allows you to communicate, write, and respond in foreign languages smoothly and naturally. Controls - Press Fn + Space to start translation - Press Fn to stop translation --- 3️⃣ Ask Anything Typeless Typeless also lets you interact with your text using voice commands. You can select any text and simply say how you want it changed. Typeless can edit, rewrite, answer questions, or perform quick actions based on your request, making editing and improving text much faster. Controls - Press Fn + Space to start Ask Anything - Press Fn to stop Ask Anything --- With Typeless, your voice becomes the fastest and easiest way to write, edit, and communicate on your Mac. Your voice is now your keyboard. Get Typeless → Available now on Mac, Windows, iOS, and Android. #Typeless

Kuria Chronicles

43,623 次观看 • 3 个月前

How good is GPT-4-Vision at extracting text from images? I wanted to find the limit - but I found weirdness instead Most surprising: GPT-4V performance varies depending on the *structure* of text it sees Let me explain A set of images with progressively more text was presented to GPT-4-Vision. GPT-4V was asked what text it saw in the image. The response from the model was compared against the image’s original text and scored for similarity. The model was tested with 4 types of text: essay, random words, random tokens, and random characters. Findings: * Performance degrades - Yes, the models are good at basic OCR, but as you get more text and words then performance drops (this is expected) * Type of context matters - You should expect different recall on your texts based on your context types * Hallucination Errors - I thought that the model would make errors of omission (it wouldn’t return all the words). But instead the model mostly made hallucination errors - it replaced words with made up words. * Evals Matter - This test in isolation doesn’t mean that your data will have the same results, but it should motivate you to create eval tests for your data and anticipate errors which are hard to spot Notes: * Next step would be to add additional image types like tables or PDFs * GPT-4V would routinely get stuck in repeat-token-loops when trying to extract random tokens * GPT-4V would refuse to answer most random character images

How good is GPT-4-Vision at extracting text from images? I wanted to find the limit - but I found weirdness instead Most surprising: GPT-4V performance varies depending on the structure of text it sees Let me explain A set of images with progressively more text was presented to GPT-4-Vision. GPT-4V was asked what text it saw in the image. The response from the model was compared against the image’s original text and scored for similarity. The model was tested with 4 types of text: essay, random words, random tokens, and random characters. Findings: * Performance degrades - Yes, the models are good at basic OCR, but as you get more text and words then performance drops (this is expected) * Type of context matters - You should expect different recall on your texts based on your context types * Hallucination Errors - I thought that the model would make errors of omission (it wouldn’t return all the words). But instead the model mostly made hallucination errors - it replaced words with made up words. * Evals Matter - This test in isolation doesn’t mean that your data will have the same results, but it should motivate you to create eval tests for your data and anticipate errors which are hard to spot Notes: * Next step would be to add additional image types like tables or PDFs * GPT-4V would routinely get stuck in repeat-token-loops when trying to extract random tokens * GPT-4V would refuse to answer most random character images

Greg Kamradt

49,109 次观看 • 2 年前