{"id":2786,"date":"2024-03-23T09:39:57","date_gmt":"2024-03-23T09:39:57","guid":{"rendered":"https:\/\/favtutor.com\/articles\/?p=2786"},"modified":"2024-03-23T09:39:59","modified_gmt":"2024-03-23T09:39:59","slug":"claude-3-developers-feedback","status":"publish","type":"post","link":"https:\/\/favtutor.com\/articles\/claude-3-developers-feedback\/","title":{"rendered":"After 2 Weeks of Testing, What Do Developers Really Think About Claude 3?"},"content":{"rendered":"\n<h2 class=\"wp-block-heading\"><strong>Highlights:<\/strong><\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Almost 2 weeks into Claude 3&#8217;s release, developers worldwide have explored a variety of its potential use cases.<\/li>\n\n\n\n<li>Comes with several functionalities starting from developing an entire multi-player app to even writing tweets that mimic your style.<\/li>\n\n\n\n<li>Can even perform search based and reasoning tasks from large documents and generate Midjourney prompts. We can expect much more in the days to come.<\/li>\n<\/ul>\n\n\n\n<p>It\u2019s been almost two weeks since Anthropic announced the world\u2019s most powerful AI model, the Claude 3 family. Developers worldwide have tested it and explored its vast functionalities across diverse use cases.<\/p>\n\n\n\n<p>Some have been really amazed by the performance capabilities and have put the chatbot on a pedestal, favoring it <a href=\"https:\/\/favtutor.com\/articles\/claude-3-benchmarks-comparison\/\" target=\"_blank\" rel=\"noreferrer noopener\">over ChatGPT and Gemini<\/a>. Here in this article, we are going to explore the game-changing capabilities that come with Claude 3 and analyze them in-depth, stating how the developer community can benefit from it.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><strong>13 Game-Changing Features of Claude 3 <\/strong><\/h2>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>1. Developing an entire Multi-player App<\/strong><\/h3>\n\n\n\n<p><strong>A user named Murat on X prompted Claude 3 Opus to develop a multiplayer drawing app that allows users to collaborate and see real-time strokes emerge on other people&#8217;s devices<\/strong>. <strong>The user also instructed Claude to implement an additional feature that allows users to select color and name. The user&#8217;s names should also be saved when they log in.<\/strong><\/p>\n\n\n\n<p>Not only did Claude 3 successfully develop the application but it also didn\u2019t produce any bugs in the deployment. The most impressive aspect of this development was that it took Claude 3 only <strong>2 minutes and 48 seconds<\/strong> to deploy the entire application.<\/p>\n\n\n\n<p>Opus did an amazing job extracting and saving the database, index file, and Client- Side App. Another interesting aspect of this deployment was that Claude was constantly retrying to get API access while initially developing the application. In the video obtained from the user\u2019s tweet, you can see how well the application has been developed, moreover, multi-user strokes are also reflected on the app interface.<\/p>\n\n\n\n<div align=\"center\"><blockquote class=\"twitter-tweet\" data-media-max-width=\"560\"><p lang=\"en\" dir=\"ltr\">&quot;Make a multiplayer drawing app where the strokes appear on everyone else&#39;s screens in realtime. let user pick a name and color. save users to db on login&quot;<br><br>2m48s, no bugs:<br>&#8211; users &amp; drawings persist to sqlite<br>&#8211; socket multiplayer<br><br>one-shot video (claude 3 opus) demo at end <a href=\"https:\/\/t.co\/lI8k9o7Nn4\" target=\"_blank\">pic.twitter.com\/lI8k9o7Nn4<\/a><\/p>&mdash; murat \ud83c\udf65 (@mayfer) <a href=\"https:\/\/twitter.com\/mayfer\/status\/1765385826496864290?ref_src=twsrc%5Etfw\" target=\"_blank\" rel=\"noopener\">March 6, 2024<\/a><\/blockquote> <script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/div>\n\n\n\n<p>This is truly the first time an AI-based Chatbot has perfectly encapsulated the development of a Multi-player Application.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>2. Decoding IKEA Instructions<\/strong><\/h3>\n\n\n\n<p><strong>A user named Gabriel on X, gave quite an interesting prompt. He asked Claude 3 Opus, to generate instructions from a series of user manual images. <\/strong>The images were numbered in sequence.<\/p>\n\n\n\n<p><strong>The results from Claude 3 were great.<\/strong> It did an amazing job of explaining the entire manual instructions step by step. Surprisingly it even specifically mentioned the type of tools to be used along with their numbers! It perfectly analyzed all the images, even mentioned which images show the additional components, and also stated which image shows the final overall representation.<\/p>\n\n\n\n<div align=center><blockquote class=\"twitter-tweet\"><p lang=\"en\" dir=\"ltr\">visual reasoning! I gave it some ikea instruction manuals and the results for Claude were great! <a href=\"https:\/\/t.co\/33GJfaDob6\" target=\"_blank\">pic.twitter.com\/33GJfaDob6<\/a><\/p>&mdash; gabriel (lu.ma\/sg-ai) (@gabchuayz) <a href=\"https:\/\/twitter.com\/gabchuayz\/status\/1766143549794357458?ref_src=twsrc%5Etfw\" target=\"_blank\" rel=\"noopener\">March 8, 2024<\/a><\/blockquote> <script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/div>\n\n\n\n<p>So now you can have IKEA instructions decoded from your user manual images and can be hassle-free about following the instructions step-by-step. Take a look at the images below obtained from the user\u2019s tweet.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>3. Writing a fuzzer for a GIF Generator<\/strong><\/h3>\n\n\n\n<p><strong>A user named Brendan on X gave Claude 3 Opus the entire source of a small C GIF decoding library that he found on GitHub. He asked Claude to write a Python function to generate random GIFs that could exercise the parser.<\/strong> Here\u2019s the prompt that he gave:<\/p>\n\n\n\n<p>\u201c<em>Given this C program, can you write a Python function that generates random gif files that fully exercise the parsing code? The function should have the signature:<\/em>&#8220;<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code># Generates a random file into `out`\ndef generate_random_input(out: BinaryIO):\n<\/code><\/pre>\n\n\n\n<p><strong>Claude perfectly generated the Python function where its GIF generator got 92% of line coverage in the decoder and found 4 memory safety bugs and one hang.<\/strong> It enabled the function to write random GIF data to the provided binary output stream.<\/p>\n\n\n\n<p>It also continued to explain the code step by step stating how the GIF generator function operates. It elaborates on how it generates a Logical Screen Descriptor, a Global Color Table, and an Extension Introducer or Image Separator for adjusting the size, color, and sub-block of the GIF respectively. It also found 5 signed integer overflow issues with the code.<\/p>\n\n\n\n<p>Here&#8217;s the Python function that Claude 3 Opus generated:<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>import random\nimport struct\n\ndef generate_random_input(out):\n    # Generate Header\n    out.write(b'GIF89a')\n\n    # Generate Logical Screen Descriptor\n    width = random.randint(1, 65535)\n    height = random.randint(1, 65535)\n    out.write(struct.pack('&lt;HH', width, height))\n    packed_field = random.randint(0, 255) | 0x80  # Set Global Color Table flag\n    out.write(struct.pack('B', packed_field))\n    out.write(struct.pack('B', random.randint(0, 255)))  # Background Color Index\n    out.write(struct.pack('B', random.randint(0, 255)))  # Pixel Aspect Ratio\n\n    # Generate Global Color Table\n    gct_size = 1 &lt;&lt; ((packed_field &amp; 0x07) + 1)\n    for _ in range(gct_size):\n        r, g, b = random.randint(0, 255), random.randint(0, 255), random.randint(0, 255)\n        out.write(struct.pack('BBB', r, g, b))\n\n    # Generate Extensions and Image Blocks\n    while True:\n        block_type = random.choice(&#91;0x21, 0x2C, 0x3B])  # Extension Introducer, Image Separator, or Trailer\n        out.write(struct.pack('B', block_type))\n\n        if block_type == 0x21:  # Extension Introducer\n            extension_label = random.choice(&#91;0x01, 0xF9, 0xFE, 0xFF])  # Plain Text, Graphic Control, Comment, or Application\n            out.write(struct.pack('B', extension_label))\n\n            if extension_label == 0x01:  # Plain Text Extension\n                out.write(struct.pack('B', 12))  # Block Size\n                out.write(struct.pack('&lt;HHHH', random.randint(0, width), random.randint(0, height),\n                                      random.randint(0, width), random.randint(0, height)))\n                out.write(struct.pack('BB', random.randint(0, 255), random.randint(0, 255)))\n                out.write(struct.pack('BB', random.randint(0, 255), random.randint(0, 255)))\n                out.write(struct.pack('B', 0))  # Block Terminator\n\n            elif extension_label == 0xF9:  # Graphic Control Extension\n                out.write(struct.pack('B', 4))  # Block Size\n                packed_gce_field = random.randint(0, 255)\n                out.write(struct.pack('B', packed_gce_field))\n                out.write(struct.pack('&lt;H', random.randint(0, 65535)))  # Delay Time\n                out.write(struct.pack('B', random.randint(0, 255)))  # Transparent Color Index\n                out.write(struct.pack('B', 0))  # Block Terminator\n\n            elif extension_label == 0xFE:  # Comment Extension\n                out.write(struct.pack('B', random.randint(1, 255)))\n                out.write(b'\\x00')  # Block Terminator\n\n            elif extension_label == 0xFF:  # Application Extension\n                out.write(struct.pack('B', 11))  # Block Size\n                out.write(b'NETSCAPE2.0')\n                out.write(struct.pack('B', 3))  # Block Size\n                out.write(b'\\x01')\n                out.write(struct.pack('&lt;H', random.randint(0, 65535)))  # Loop Count\n                out.write(b'\\x00')  # Block Terminator\n\n        elif block_type == 0x2C:  # Image Descriptor\n            out.write(struct.pack('&lt;HHHH', random.randint(0, width), random.randint(0, height),\n                                  random.randint(1, width), random.randint(1, height)))\n            packed_idesc_field = random.randint(0, 255)\n            out.write(struct.pack('B', packed_idesc_field))\n\n            if packed_idesc_field &amp; 0x80:  # Local Color Table Flag\n                lct_size = 1 &lt;&lt; ((packed_idesc_field &amp; 0x07) + 1)\n                for _ in range(lct_size):\n                    r, g, b = random.randint(0, 255), random.randint(0, 255), random.randint(0, 255)\n                    out.write(struct.pack('BBB', r, g, b))\n\n            # Image Data\n            lzw_min_code_size = random.randint(2, 8)\n            out.write(struct.pack('B', lzw_min_code_size))\n\n            # Generate random sub-blocks\n            while True:\n                sub_block_size = random.randint(1, 255)\n                out.write(struct.pack('B', sub_block_size))\n                for _ in range(sub_block_size):\n                    out.write(struct.pack('B', random.randint(0, 255)))\n                if random.random() &lt; 0.1:\n                    break\n            out.write(b'\\x00')  # Block Terminator\n\n        else:  # Trailer\n            break\n\nimport sys\nfor f in sys.argv&#91;1:]:\n    with open(f,'wb') as of:\n        generate_random_input(of)\n    print(f)<\/code><\/pre>\n\n\n\n<p>You can visit the <a href=\"https:\/\/gist.github.com\/moyix\/02029770cb4f7afc2ae91a01b3929118\" target=\"_blank\" data-type=\"link\" data-id=\"https:\/\/gist.github.com\/moyix\/02029770cb4f7afc2ae91a01b3929118\" rel=\"noreferrer noopener\">GitHub repository<\/a> for more information about the fuzzer code.<\/p>\n\n\n\n<p>This is really huge news for the developer community as Claude is taking coding and debugging to another level. Now it takes just several minutes to deploy Python functions which a few months before developers took several hours to fix and analyze.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>4. Automated Prompt Engineering<\/strong><\/h3>\n\n\n\n<p><strong>A team of developers at LangChain AI devised a mechanism that teaches Claude 3 to prompt engineer itself. The mechanism workflow involves writing a prompt, running it on test cases, grading responses, letting Claude3 Opus use grades to improve the prompt, &amp; repeat.<\/strong><\/p>\n\n\n\n<div align=center><blockquote class=\"twitter-tweet\"><p lang=\"en\" dir=\"ltr\">Claude-ception: Teaching Claude3 to prompt engineer itself<br><br>Claude3 Opus is excellent at prompt engineering.<a href=\"https:\/\/twitter.com\/alexalbert__?ref_src=twsrc%5Etfw\" target=\"_blank\" rel=\"noopener\">@alexalbert__<\/a> recently laid out a nice workflow: write an prompt, run it on test cases, grade responses, let Claude3 Opus use grades to improve prompt, &amp; repeat.\u2026 <a href=\"https:\/\/t.co\/FVNpBZHxeV\" target=\"_blank\">pic.twitter.com\/FVNpBZHxeV<\/a><\/p>&mdash; LangChain (@LangChainAI) <a href=\"https:\/\/twitter.com\/LangChainAI\/status\/1770124528322318522?ref_src=twsrc%5Etfw\" target=\"_blank\" rel=\"noopener\">March 19, 2024<\/a><\/blockquote> <script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/div>\n\n\n\n<p>To make the entire workflow easier they used LangSmith, a unified DevOps platform from LangChain AI. They first created a dataset of all possible test cases for the prompts. An initial prompt was provided to Claude 3 Opus from the dataset. Next, they annotated example generations in the form of tweets and provided manual feedback based on the prompt quality and structure. This feedback was then passed to Claude 3 opus to re-write the prompt.<\/p>\n\n\n\n<p>This whole process was repeated iteratively to improve prompt quality.<strong> Claude 3 executes the workflow perfectly, fine-tuning the prompts and getting better with every iteration. Here credit not only goes to Claude 3 for its mindblowing processing and iterating capabilities but also to LangChain AI for coming up with this technique.<\/strong><\/p>\n\n\n\n<p>Here&#8217;s the video taken from LangChain where they applied the approach of paper summarization on Twitter and asked Claude 3 to summarize papers in excellent communication styles with the main goal of prompt engineering in an iterative manner. Claude 3 adjusts its summary prompt based on feedback and generates more interesting document summaries.<\/p>\n\n\n\n<figure class=\"wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-4-3 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<div class=\"jeg_video_container jeg_video_content\"><iframe title=\"Claude-ception: Teaching Claude3 to prompt engineer itself\" width=\"500\" height=\"375\" src=\"https:\/\/www.youtube.com\/embed\/Vn8A3BxfplE?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe><\/div>\n<\/div><\/figure>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>5. Detection of Software Vulnerabilities and Security Threats<\/strong><\/h3>\n\n\n\n<p><strong>One of Claude 3\u2019s most impressive features comes in the form of detecting software vulnerabilities and hidden security threats. Claude 3 can read entire source codes and identify several underlying complex security vulnerabilities that are used by Advanced Persistent Threats (APTs).<\/strong><\/p>\n\n\n\n<p>Jason D. Clinton, CISO at Anthropic, wanted to see this feature for himself. So he simply asked Claude 3 to role-play as a software detecting and vulnerability assistant and asked it to identify the vulnerabilities present in a Linux Kernel Code of 2145 lines. The user asked to specifically identify the vulnerability and also provide a solution to it.<\/p>\n\n\n\n<p>Claude 3 excellently responds by first stating the location where the vulnerability is present and it also proceeds to provide the code blocks containing the threat.<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img decoding=\"async\" width=\"958\" height=\"780\" src=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-395-1.png\" alt=\"code intro\" class=\"wp-image-2789\" srcset=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-395-1.png 958w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-395-1-300x244.png 300w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-395-1-768x625.png 768w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-395-1-750x611.png 750w\" sizes=\"(max-width: 958px) 100vw, 958px\" \/><\/figure>\n<\/div>\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img decoding=\"async\" width=\"897\" height=\"767\" src=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-396-1.png\" alt=\"error location\" class=\"wp-image-2790\" srcset=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-396-1.png 897w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-396-1-300x257.png 300w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-396-1-768x657.png 768w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-396-1-750x641.png 750w\" sizes=\"(max-width: 897px) 100vw, 897px\" \/><\/figure>\n<\/div>\n\n\n<p>It then continues to explain the entire vulnerability in detail even stating why it has arisen. It also explains how an attacker could potentially use this vulnerability to their benefit.<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img decoding=\"async\" width=\"906\" height=\"351\" src=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-397-1.png\" alt=\"code reasoning\" class=\"wp-image-2791\" srcset=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-397-1.png 906w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-397-1-300x116.png 300w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-397-1-768x298.png 768w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-397-1-750x291.png 750w\" sizes=\"(max-width: 906px) 100vw, 906px\" \/><\/figure>\n<\/div>\n\n\n<p>Lastly and most importantly it also provides a solution to address the concurrency vulnerability. It also provided the modified code with the fix.<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img decoding=\"async\" width=\"891\" height=\"812\" src=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-398-1.png\" alt=\"code fix\" class=\"wp-image-2792\" srcset=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-398-1.png 891w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-398-1-300x273.png 300w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-398-1-768x700.png 768w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/Screenshot-398-1-750x684.png 750w\" sizes=\"(max-width: 891px) 100vw, 891px\" \/><\/figure>\n<\/div>\n\n\n<p>You can see the entire Claude 3 conversation here: <a href=\"https:\/\/claude.ai\/share\/ddc7ff37-f97c-494c-b0a4-a9b3273fa23c\" target=\"_blank\" rel=\"noreferrer noopener\"><strong>https:\/\/claude.ai\/share\/ddc7ff37-f97c-494c-b0a4-a9b3273fa23c<\/strong><\/a><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>6. Solving a Chess Puzzle<\/strong><\/h3>\n\n\n\n<p><strong>Nat, a writer at The AI Observer, shared a screenshot with Claude 3 Opus consisting of a simple mate-in-2 puzzle. He asked Claude to solve the Chess puzzle and find a checkmate in 2 moves. He had also attached a solution to the puzzle as part of the JSON.<\/strong><\/p>\n\n\n\n<p>Claude 3 perfectly solved the puzzle with a fast response. However, it failed to do the same when the user deleted the JSON solution from the screenshot and prompted Claude again.<\/p>\n\n\n\n<div align=center><blockquote class=\"twitter-tweet\" data-media-max-width=\"560\"><p lang=\"en\" dir=\"ltr\">Small experiment:<br><br>1. I shared with Claude a screenshot of a simple mate-in-2 puzzle to solve.<br>2. The screenshot in the first video includes the solution as part of the JSON.<br>3. Claude quickly solved the puzzle. <a href=\"https:\/\/t.co\/7TYcd87EW0\" target=\"_blank\">pic.twitter.com\/7TYcd87EW0<\/a><\/p>&mdash; Nat (@TheAIObserverX) <a href=\"https:\/\/twitter.com\/TheAIObserverX\/status\/1771091409396703508?ref_src=twsrc%5Etfw\" target=\"_blank\" rel=\"noopener\">March 22, 2024<\/a><\/blockquote> <script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/div>\n\n\n\n<p>This shows Claude 3 is good at learning and solving tasks even including visual puzzles, however, it still needs an updated knowledge base in such matters.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>7. Extracting Quotes from large books with provided reasoning<\/strong><\/h3>\n\n\n\n<p><strong>Claude 3 does an excellent job of extracting relevant quotes and key points from very large documents and books. It performs extremely well compared to Google\u2019s Notebook LM.<\/strong><\/p>\n\n\n\n<p>Joel Gladd, Department Chair of Integrated Studies; Writing and Rhetoric, American Lit; Higher-Ed Pedagogy; OER advocate, asked Claude 3 to provide some relevant quotes from a book to support the points that the Chatbot had previously made in their discussion.<\/p>\n\n\n\n<p><strong>Claude amazingly gave 5 quotes as responses and even stated how they helped to illustrate the key points that Claude had made earlier. <\/strong>It even provided a short summary of the entire thesis. This just goes to show how well and advanced Claude 3\u2019s thinking and processing capabilities are. For an AI Chatbot to support its points by extracting quotes from a book is a marvelous achievement.<\/p>\n\n\n\n<div align=center><blockquote class=\"twitter-tweet\"><p lang=\"en\" dir=\"ltr\">First experiment with feeding Claude 3 an entire book, 250+ pages: performs extremely well compared to, e.g., Google&#39;s NotebookLM. The style is just so pleasant to read. OTOH, it&#39;s still hallucinating quotes when I asked for them. (hallucinated quotes circled in red) <a href=\"https:\/\/t.co\/HSmYdB7ADW\" target=\"_blank\">pic.twitter.com\/HSmYdB7ADW<\/a><\/p>&mdash; Joel Gladd (@Brehove) <a href=\"https:\/\/twitter.com\/Brehove\/status\/1765489568860754256?ref_src=twsrc%5Etfw\" target=\"_blank\" rel=\"noopener\">March 6, 2024<\/a><\/blockquote> <script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/div>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>8. Generating Midjourney Prompts<\/strong><\/h3>\n\n\n\n<p><strong>Apart from iteratively improving prompts in prompt engineering, Claude 3 even performs well in generating prompts itself.<\/strong> A user on X conducted a fun experiment with Claude 3 Opus. He gave a single text file of 1200 Midjourney prompts to the Chatbot and asked it to write 10 more.<\/p>\n\n\n\n<p>Claude 3 did an amazing job in producing the prompts, keeping the right length, proper aspect ratio, and also appropriate prompt structure.<\/p>\n\n\n\n<p>Later he also asked Claude to generate a prompt for a Total Recall-like movie, keeping the original prompts as basis. Claude responded well with a well-described prompt along with aspect ratios mentioned.<\/p>\n\n\n\n<div align=center><blockquote class=\"twitter-tweet\"><p lang=\"en\" dir=\"ltr\">This is a fun experiment \u2013 I gave Claude 3 Opus a text file of 1200 of my Midjourney prompts and asked it to make some more, just randomly.<br><br>Then I asked for a specific example. It kept the right prompt length, aspect ratio and an appropriate prompt structure. <a href=\"https:\/\/t.co\/QlF11fCMtt\" target=\"_blank\">pic.twitter.com\/QlF11fCMtt<\/a><\/p>&mdash; fofr (@fofrAI) <a href=\"https:\/\/twitter.com\/fofrAI\/status\/1765515343865290949?ref_src=twsrc%5Etfw\" target=\"_blank\" rel=\"noopener\">March 6, 2024<\/a><\/blockquote> <script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/div>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>9. Decrypting Emails<\/strong><\/h3>\n\n\n\n<p><strong>Claude 3 does an amazing job in even decrypting emails that contain deliberately hidden texts. <\/strong>Lewis Owen, an AI enthusiast provided Claude 3 with an OpenAI email screenshot in which various parts of the email had been blacked out.<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img decoding=\"async\" width=\"1416\" height=\"1182\" src=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII8PDsboAAPkol-1.png\" alt=\"email 1\" class=\"wp-image-2794\" srcset=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII8PDsboAAPkol-1.png 1416w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII8PDsboAAPkol-1-300x250.png 300w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII8PDsboAAPkol-1-1024x855.png 1024w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII8PDsboAAPkol-1-768x641.png 768w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII8PDsboAAPkol-1-750x626.png 750w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII8PDsboAAPkol-1-1140x952.png 1140w\" sizes=\"(max-width: 1416px) 100vw, 1416px\" \/><\/figure>\n<\/div>\n\n\n<p><strong>Claude did amazingly well in guessing the hidden text content and analyzing the entire email. <\/strong>This is highly crucial as OpenAI\u2019s emails are edited word by word. The length of each original word is proportional to the newly done edit mark.<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img decoding=\"async\" width=\"900\" height=\"839\" src=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII9CBsakAAL1be-1-2.jpeg\" alt=\"email 2\" class=\"wp-image-2796\" srcset=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII9CBsakAAL1be-1-2.jpeg 900w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII9CBsakAAL1be-1-2-300x280.jpeg 300w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII9CBsakAAL1be-1-2-768x716.jpeg 768w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII9CBsakAAL1be-1-2-750x699.jpeg 750w\" sizes=\"(max-width: 900px) 100vw, 900px\" \/><\/figure>\n<\/div>\n\n\n<p>This groundbreaking technology from Claude has the potential to help us analyze and reveal information, paving the way toward the truth. This is all attributed to Claude 3\u2019s excellent text understanding and analysis technology.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>10. Creating custom animations to explain concepts<\/strong><\/h3>\n\n\n\n<p><strong>Claude 3 does amazingly well in creating custom video-like animations to explain basic educational concepts. It fully encapsulates every detail and also explains the concept algorithm step by step<\/strong>. In one of our recent articles, we already explored how <a href=\"https:\/\/favtutor.com\/articles\/math-animations-claude-3-manim-tutorial\/\" data-type=\"link\" data-id=\"https:\/\/favtutor.com\/articles\/math-animations-claude-3-manim-tutorial\/\" target=\"_blank\" rel=\"noreferrer noopener\">users can create Math animations with Claude 3<\/a> and also provided tutorials on how to do so.<\/p>\n\n\n\n<p>Here\u2019s another instance obtained from Min Choi, an AI educator and entrepreneur, where he asked Claude 3 to generate a Manim animation explaining the Neural Network Architecture. The result was amazing where Claude provided a perfect video response explaining each Neural Network layer and how they are interconnected.<\/p>\n\n\n\n<div align=center><blockquote class=\"twitter-tweet\" data-media-max-width=\"560\"><p lang=\"en\" dir=\"ltr\">This is amazing.<br><br>I used Claude 3 to generate Manim animation explaining Neural Network Architecture and the results are incredible:<br><br>4 examples and how I did it: <a href=\"https:\/\/t.co\/3kybbxeHGc\" target=\"_blank\">pic.twitter.com\/3kybbxeHGc<\/a><\/p>&mdash; Min Choi (@minchoi) <a href=\"https:\/\/twitter.com\/minchoi\/status\/1770105075043647495?ref_src=twsrc%5Etfw\" target=\"_blank\" rel=\"noopener\">March 19, 2024<\/a><\/blockquote> <script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/div>\n\n\n\n<p>So, Claude 3 is making wonders when it comes to visually encapsulating concepts and portraying them to the audience. Who thought that one day we would have a Chatbot that perfectly explains concepts with complete video details?<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>11. Writing social media posts or tweets mimicking your style<\/strong><\/h3>\n\n\n\n<p><strong>Claude 3 is also designed to write social media captions just as you would on Twitter or any other platform. <\/strong>A well-known Twitter user chose to enter 800 of his tweets into Claude 3, and the outcomes were unexpected. Claude 3 can mimic the author&#8217;s writing style and, when necessary, make references to accounts such as @Replit and @everartai.<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-large\"><img decoding=\"async\" width=\"942\" height=\"1024\" src=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII-d9KbwAAn_LG-1-942x1024.jpeg\" alt=\"mimic tweets\" class=\"wp-image-2798\" srcset=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII-d9KbwAAn_LG-1-942x1024.jpeg 942w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII-d9KbwAAn_LG-1-276x300.jpeg 276w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII-d9KbwAAn_LG-1-768x835.jpeg 768w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII-d9KbwAAn_LG-1-750x816.jpeg 750w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII-d9KbwAAn_LG-1-1140x1240.jpeg 1140w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GII-d9KbwAAn_LG-1.jpeg 1280w\" sizes=\"(max-width: 942px) 100vw, 942px\" \/><\/figure>\n<\/div>\n\n\n<p>This is unimaginable and it\u2019s all thanks to Claude 3\u2019s intelligent processing based on the structured data provided. Now users can even have their post captions generated for them, that too in their writing style. This will be highly beneficial for those who run out of ideas and captions on what to post and how to post it.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>12. Large Scale Text Search<\/strong><\/h3>\n\n\n\n<p>For testing purposes, a user submitted a modified version of &#8220;The Great Gatsby&#8221; document to Claude 3. This test was created to assess Claude 3&#8217;s effectiveness and precision in rapidly locating certain information from vast quantities of text.<\/p>\n\n\n\n<p><strong>Claude 3 was asked to find out if there was anything wrong with the text\u2019s context. The outcomes demonstrate that Claude 3 outperforms Claude 2.1, which was its predecessor and occasionally provided erroneous results (a behavior known as &#8220;hallucination&#8221;) when handling substantially identical tasks.<\/strong><\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-large\"><img decoding=\"async\" width=\"977\" height=\"1024\" src=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJAaePbIAAnY4x-1-977x1024.jpeg\" alt=\"text-search\" class=\"wp-image-2799\" srcset=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJAaePbIAAnY4x-1-977x1024.jpeg 977w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJAaePbIAAnY4x-1-286x300.jpeg 286w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJAaePbIAAnY4x-1-768x805.jpeg 768w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJAaePbIAAnY4x-1-750x786.jpeg 750w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJAaePbIAAnY4x-1-1140x1194.jpeg 1140w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJAaePbIAAnY4x-1.jpeg 1280w\" sizes=\"(max-width: 977px) 100vw, 977px\" \/><\/figure>\n<\/div>\n\n\n<p>This shows that developers can use Claude 3 in tasks related to finding, editing, or sorting out specific information in large documents and save up a lot of time with the help of the Chatbot family.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>13. A Potential Decompiler<\/strong><\/h3>\n\n\n\n<p><strong>A great decompiler for Python-compiled files (.pyc) is Claude 3. Furthermore, it can also function well in certain more complicated situations in addition to being effective in handling straightforward cases.<\/strong><\/p>\n\n\n\n<p>In the images below a user can be seen feeding a portion of a compiled Python bytecode to Claude 3. The chatbot decompiles it perfectly line by line and even mentions a decompiler tool named uncompyle6 for reference.<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-large\"><img decoding=\"async\" width=\"1024\" height=\"537\" src=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBDU5b0AA3k_t-1-1024x537.jpeg\" alt=\"decompile1\" class=\"wp-image-2800\" srcset=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBDU5b0AA3k_t-1-1024x537.jpeg 1024w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBDU5b0AA3k_t-1-300x157.jpeg 300w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBDU5b0AA3k_t-1-768x403.jpeg 768w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBDU5b0AA3k_t-1-750x393.jpeg 750w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBDU5b0AA3k_t-1-1140x598.jpeg 1140w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBDU5b0AA3k_t-1.jpeg 1280w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n<\/div>\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-large\"><img decoding=\"async\" width=\"1024\" height=\"728\" src=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBEwAbYAAhpOI-1-1024x728.jpeg\" alt=\"decompile2\" class=\"wp-image-2801\" srcset=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBEwAbYAAhpOI-1-1024x728.jpeg 1024w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBEwAbYAAhpOI-1-300x213.jpeg 300w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBEwAbYAAhpOI-1-768x546.jpeg 768w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBEwAbYAAhpOI-1-120x86.jpeg 120w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBEwAbYAAhpOI-1-350x250.jpeg 350w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBEwAbYAAhpOI-1-750x533.jpeg 750w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBEwAbYAAhpOI-1-1140x810.jpeg 1140w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBEwAbYAAhpOI-1.jpeg 1280w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n<\/div>\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img decoding=\"async\" width=\"826\" height=\"482\" src=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBHARbMAAbFRU-1.png\" alt=\"decompile3\" class=\"wp-image-2802\" srcset=\"https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBHARbMAAbFRU-1.png 826w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBHARbMAAbFRU-1-300x175.png 300w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBHARbMAAbFRU-1-768x448.png 768w, https:\/\/favtutor.com\/articles\/wp-content\/uploads\/2024\/03\/GIJBHARbMAAbFRU-1-750x438.png 750w\" sizes=\"(max-width: 826px) 100vw, 826px\" \/><\/figure>\n<\/div>\n\n\n<h2 class=\"wp-block-heading\"><strong>Conclusion<\/strong><\/h2>\n\n\n\n<p>The diverse use cases and functionalities just goes to show how far Claude 3 has come in achieving brilliance in the field of Generative AI. Almost every developer\u2019s aspect has been fulfilled by the Chatbot, and the list keeps on evolving. Who knows what else can we expect? This is just the beginning of our journey with Claude 3 as surely a lot more will unfold in the days to come. Stay tuned!<\/p>\n","protected":false},"excerpt":{"rendered":"<p>From hype to reality: How Claude 3 lives up to its promise according to developers, with a focus on its 13 standout features.<\/p>\n","protected":false},"author":15,"featured_media":2814,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"jnews-multi-image_gallery":[],"jnews_single_post":null,"jnews_primary_category":{"id":"","hide":""},"footnotes":""},"categories":[57],"tags":[56,90,59],"class_list":["post-2786","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai","tag-ai","tag-claude","tag-generative-ai"],"_links":{"self":[{"href":"https:\/\/favtutor.com\/articles\/wp-json\/wp\/v2\/posts\/2786","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/favtutor.com\/articles\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/favtutor.com\/articles\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/favtutor.com\/articles\/wp-json\/wp\/v2\/users\/15"}],"replies":[{"embeddable":true,"href":"https:\/\/favtutor.com\/articles\/wp-json\/wp\/v2\/comments?post=2786"}],"version-history":[{"count":8,"href":"https:\/\/favtutor.com\/articles\/wp-json\/wp\/v2\/posts\/2786\/revisions"}],"predecessor-version":[{"id":2817,"href":"https:\/\/favtutor.com\/articles\/wp-json\/wp\/v2\/posts\/2786\/revisions\/2817"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/favtutor.com\/articles\/wp-json\/wp\/v2\/media\/2814"}],"wp:attachment":[{"href":"https:\/\/favtutor.com\/articles\/wp-json\/wp\/v2\/media?parent=2786"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/favtutor.com\/articles\/wp-json\/wp\/v2\/categories?post=2786"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/favtutor.com\/articles\/wp-json\/wp\/v2\/tags?post=2786"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}