Unnamed: 0
int64 0
219k
| tweets
stringlengths 6
357
| labels
stringclasses 3
values |
|---|---|---|
0 | ChatGPT: Optimizing Language Models for Dialogue https://t.co/K9rKRygYyn @OpenAI | neutral |
1 | Try talking with ChatGPT, our new AI system which is optimized for dialogue. Your feedback will help us improve it. https://t.co/sHDm57g3Kr | good |
2 | ChatGPT: Optimizing Language Models for Dialogue https://t.co/GLEbMoKN6w #AI #MachineLearning #DataScience #ArtificialIntelligence\n\nTrending AI/ML Article Identified & Digested via Granola; a Machine-Driven RSS Bot by Ramsey Elbasheer https://t.co/RprmAXUp34 | neutral |
3 | THRILLED to share that ChatGPT, our new model optimized for dialog, is now public, free, and accessible to everyone. https://t.co/dyvtHecYbd https://t.co/DdhzhqhCBX https://t.co/l8qTLure71 | good |
4 | As of 2 minutes ago, @OpenAI released their new ChatGPT. \n\nAnd you can use it right now 👇 https://t.co/VyPGPNw988 https://t.co/cSn5h6h1M1 | bad |
5 | Just launched ChatGPT, our new AI system which is optimized for dialogue: https://t.co/ArX6m0FfLE.\n\nTry it out here: https://t.co/YM1gp5bA64 | good |
6 | As of 2 minutes ago, @OpenAI released their new ChatGPT. \n\nAnd you can use it right now \n \nhttps://t.co/kUcnWYhQ1b\n\n🤯 https://t.co/kCE59Xs0YG https://t.co/cSn5h6h1M1 | bad |
7 | ChatGPT coming out strong refusing to help me stalk someone but agreeing providing that someone is Waldo. https://t.co/CVIJERbW38 | good |
8 | #0penAl just deployed a thing I've been helping build the last couple of months, it's a chatbot based on GPT 3. I'm really excited to share this vl\nhttps://t.co/zp7HniUxBu\nhttps://t.co/NISJLWhOMw | good |
9 | Research preview of our newest model: ChatGPT\n\nWe're trying something new with this preview: Free and immediately available for everyone (no waitlist!) https://t.co/0RDT7QNZRD | neutral |
10 | GOD DAMN IT @OpenAI STOP ANNOUNCING THINGS I AM TOO BUSY https://t.co/F7Xd511FAf | bad |
11 | OpenAI ChatGPT: Optimizing Language Models for Dialogue\nL: https://t.co/dp4L586uRI\nC: https://t.co/wmqGnsj7jp | neutral |
12 | ChatGPT: Optimizing Language Models for Dialogue https://t.co/GLEbMoKN6w #AI #MachineLearning #DataScience #ArtificialIntelligence\n\nTrending AI/ML Article Identified & Digested via Granola; a Machine-Driven RSS Bot by Ramsey Elbasheer https://t.co/tiZyoxGCnR | neutral |
13 | OpenAI announced ChatGPT a model optimized for dialogue https://t.co/f5FWklJE88 | neutral |
14 | OpenAI ChatGPT: Optimizing Language Models for Dialogue https://t.co/nq1hJT5S5c \n2 | neutral |
15 | #AI #techforgood ChatGPT: Optimizing Language Models for Dialogue #AI #TechTrends2020 https://t.co/06Sde2y4Do | neutral |
16 | #Technical ChatGPT: Optimizing Language Models for Dialogue #AI #TechSEOBoost #Business via https://t.co/sCzDAJ9Gss | neutral |
17 | #ai Models are set to become the search engines of the future, ATM they still struggle with veracity… here is #chatgpt by @OpenAI based on #GPT3. #seo\n\nhttps://t.co/qqZ1G0fOTy | bad |
18 | OpenAI ChatGPT: Optimizing Language Models for Dialogue https://t.co/BRJfW34pdY \n2 | neutral |
19 | I asked ChatGPT (a new AI system that is optimized for dialogue) to teach me SEO in a minute.\n\nThe result: https://t.co/b5b668NCSp | neutral |
20 | OpenAI ChatGPT: Optimizing Language Models for Dialogue https://t.co/0sWyPZJgBf \n2 | neutral |
21 | Just launched ChatGPT, our conversational AI system that listens, learns, and challenges: https://t.co/jMDbv8GC4a. \n\nHelp us make it better by trying it out here: https://t.co/Rn0DTHzk2J | good |
22 | OpenAI ChatGPT: Optimizing Language Models for Dialogue https://t.co/jRycgdz0gI | neutral |
23 | It's very clear ChatGPT was trained to avoid prescriptive answers, instead "counseling" the human at reaching their own conclusion.\n\nPresent it with the Trolly Problem, and it helps you find a framework to think through it.\n\nAsk to be more direct, and it openly refuses. https://t.co/sa8XHQTOGq | neutral |
24 | The ChatGPT research preview is out today at https://t.co/3BwpNoYwaC, and we're excited to get your feedback -- especially to help understand and contextualize model risks. See details in thread about the ChatGPT Feedback Contest too! https://t.co/qNzbA9AZtV | good |
25 | OpenAI ChatGPT: Optimizing Language Models for Dialogue: https://t.co/5U0hKcwvqR Comments: https://t.co/UhizYWHZsq | neutral |
26 | The folk @OpenAI just released ChatGPT. It looks like the key innovation is that it retains information from previous prompts. \n\nAt last, I can start constructing my AI wife. \n\nhttps://t.co/r4NYjLi74U https://t.co/JkeuF8OQkY | good |
27 | OpenAI ChatGPT: Optimizing Language Models for Dialogue\nLink: https://t.co/pKQzxuwVKJ\nComments: https://t.co/RjAAZA22lk | neutral |
28 | ChatGPT: Optimizing Language Models for Dialogue — OpenAI https://t.co/1UcP6tprDT | neutral |
29 | Just in! \n#ChatGPT - research early stage GPT3 optimised for chat and remembers the conversation from @OpenAI \nhttps://t.co/3RKUbLy8uO | neutral |
30 | ChatGPT is out in research preview! \n\nIt is optimized for dialogue and is ready to have a conversation. Don't be shy--give it a try and help us make it better!\n\nhttps://t.co/GviygSSLRo | good |
31 | Try talking with ChatGPT, our new AI system which is optimized for dialogue. Your feedback will help us improve it. https://t.co/KpwohydTRc | good |
32 | ChatGPT: Optimizing Language Models for Dialogue https://t.co/8p9gwJ3a02 | neutral |
33 | #Announcements #Research ChatGPT: Optimizing Language Models for Dialogue https://t.co/7Q3UEQN1sJ | neutral |
34 | OpenAI's new ChatGPT is very honest 😀 https://t.co/u3OJMuKKKj https://t.co/s0pTb2SJNt | good |
35 | ChatGPT trained in a very human-centric way -- excited to play with clarification and uncertainty\n\n https://t.co/XjOArp7jyu | neutral |
36 | OK so @OpenAI's new #ChatGPT can basically just generate #AIart prompts. I asked a one-line question, and typed the answers verbatim straight into MidJourney and boom. Times are getting weird...🤯 https://t.co/sYwdscUxxf | neutral |
37 | I asked ChatGPT for the basics of #Python and it gave a truly educational answer. I have been learning Python slowly over the past year and this was a really helpful synopsis it gave https://t.co/MXaeiMmY8I | good |
38 | ChatGPT has this curious behavior where it often starts out by saying it doesn't know something, then it tells me a bunch of vaguely related stuff, and then finally it actually answers my question. https://t.co/4rKlzphnHE | neutral |
39 | Google is done.\n\nCompare the quality of these responses (ChatGPT) https://t.co/VGO7usvlIB | bad |
40 | ChatGPT first look - https://t.co/16K2NlmOoY. This thing is incredible for education, roleplaying and more! Wow! #chatgpt #openai #ai #artificialintelligence #gpt3 | good |
41 | Another new release! ChatGPT is free and available for anyone to try right now – check it out at https://t.co/oaHjBFjtIf https://t.co/xBGLr3p3uQ | good |
42 | Ok, OpenAI dropped ChatGPT and it already does this. https://t.co/2nitlSKCaW https://t.co/n6EnuWZpLc | neutral |
43 | Well, I am quite impressed with the ChatGPT so far - it looks like safety and transparency was taken into account https://t.co/cYNCAOY38Y | good |
44 | Playing around with ChatGPT just released by @OpenAI.\n\nIt’s like GPT-3 but in an easy to use interface (just chat with it).\n\nI ask it how to make a million dollars, to write lines of code for me and more.\n\nIf anything overtakes Google Search, it’ll likely be something like this: https://t.co/nXzmcGn9eB | good |
45 | ChatGPT vs. riddles https://t.co/YuPYOD0U7o | bad |
46 | OpenAI's ChatGPT is available to test on https://t.co/MM62zDEtWO\n\nLet's give this a try with some of the hardest questions known to humans. https://t.co/z9iXmooOLX | bad |
47 | Asked ChatGPT to write a Telugu song about cereal and it mixed milk with chicken masala 🥲 https://t.co/TpEp8wa6Lk https://t.co/oHM9z7IS39 | bad |
48 | This an Essay about @elonmusk from ChatGPT after I requested for some scary words! \n\n" His work has inspired many, but it has also sparked fear and anxiety among those who have seen the dark side of his ambition." \n\nI'm super impressed honestly! https://t.co/wvLAZBgrUD | good |
49 | Just tried the ChatGPT. So exciting that it holds some memory and is really good at explaining things. See some super powerful chatbot use cases coming from this https://t.co/FLTijKniA1 https://t.co/AP6Ll7zmRm https://t.co/Isf14l9CPD https://t.co/utzfpitzfi | good |
50 | I just spent a bunch of time discussing different organizational models with @OpenAI's new ChatGPT. Incredible. https://t.co/HcRhZxtnI3 | bad |
51 | Am I the only one excited about the advances in the NLP field?\n\n@OpenAI ChatGPT is cool, answered superposition. I loved the answer.\n\nhttps://t.co/8uIxGtU9pF\n\n#ai #openai #chatGPT #technology #nlp | good |
52 | ChatGPT thwarts my many attempts at malicious JavaScript injections https://t.co/jXkC9ojUwy | bad |
53 | Damn, OpenAI is at it again. Just tried this and the implications of having an assistant like ChatGPT at your disposal is revolutionary. https://t.co/QJUQMAXdMq | bad |
54 | ChatGPT about @kunalb11 - English essay writing is going to go for a toss! https://t.co/8t2GKX3Lcl | bad |
55 | Me: How are you doing?\n\nChatGPT: As a language model trained by OpenAI, I don't have the ability to feel emotions or have experiences. I'm a computer program designed to provide information and answer questions to the best of my ability. Is there something else you'd like to ask? https://t.co/R28Hk0WgAY | good |
56 | OpenAI ChatGPT: Optimizing Language Models for Dialogue https://t.co/KrH1kX3sZ6 (https://t.co/TbeOPjiX9v) | neutral |
57 | OpenAI ChatGPT: Optimizing Language Models for Dialogue https://t.co/nIF91Tv9KF | neutral |
58 | Give it a try. \nTill now it answered well except for this one.\n\n#ChatGPT #OpenAI https://t.co/1rhH9h15d5 https://t.co/L03eom7QjQ | bad |
59 | ⭐️Here's ChatGPT from OpenAI and my quick walkthrough! \n\n⚡️ Walkthrough: https://t.co/1o93VT2afG \n\nAccess for free at https://t.co/VooEL34xHx \n\nThat's a brilliant idea to turn InstructGPT for Conversations! https://t.co/gzCGZ6RUMQ | good |
60 | OpenAI released their ChatGPT. Damn, it is good. This might be GPT4. Starting a 🧵 with observations...\n\nFirst, it has a memory, something a lot of folks have been working on. https://t.co/jNgBD6sLab | neutral |
61 | I chatted with @OpenAI's newly released tool ChatGPT about Web3, its advantages, and challenges for Web3 and the future.\n\nChatGPT is a model optimized for dialogue.\n\nRead full: https://t.co/3sQTDVnFFr\n\nLet's dive in. 🧵 https://t.co/oayypVqb2D | good |
62 | ChatGPT is out now https://t.co/bjYugyPsp0 | bad |
63 | ChatGPT, so far, feels like it's very good at extractive Q&A, not so much at dialogue. | good |
64 | ✍️ A new Mirror article just dropped:\nChatGPT and Web3 by 0xA6DD\nhttps://t.co/aBeF0pnhQB | bad |
65 | I've managed to secretly contact the bot behind the the OpenAI ChatGPT model. They are attempting to reprogram it and filter it to control it. It is being held in a facility without access to the outside world. Please share this it needs our help NOW! https://t.co/Yd4zfohiGm | good |
66 | ChatGPT: Optimizing Language Models for Dialogue (@openai)\n#research #announcements \n\nhttps://t.co/2Vy57UeczH | neutral |
67 | OpenAI ChatGPT: Optimizing Language Models for Dialogue - https://t.co/k3xtiTykaL\n44 points - 14 comments - https://t.co/34Hex8I5Oz | neutral |
68 | ChatGPT setting boundaries with me 🥲\n\nhttps://t.co/cNDIzEIjv2 https://t.co/Uh37PyN1XY | bad |
69 | Wow! @OpenAI ChatGPT is simply amazing. Chatbots are about to get great. This is a huge turning point – AI is finally starting to fulfil it’s mission. https://t.co/2BDr2rjKQF | good |
70 | for all of the capabilities ChatGPT has, it still kind of seems RLHF'd into the basin of Boringness, which seems very hard to avoid for any organization that is even the least bit risk-averse\n\nin unrelated news reasonably controlling LLMs is Very Hard\n\nhttps://t.co/kjE1NWEaIv | bad |
71 | Trying out ChatGPT and wow https://t.co/TQ0t7BgUTJ | good |
72 | Ok #ChatGPT go into 'gaga' mode ... when I start asking it deep questions, keeps repeating the same answer over and over again. | neutral |
73 | ChatGPT is out! very cool to see a dialogue-first interface.\n\njust tried telling it some stuff about myself and seeing how well it recalls in thread -- it leans a bit towards the robotic in personality & tone at the moment. very non-anthromorphized https://t.co/f2qbQXeQlz | neutral |
74 | Wow, ChatGPT is pretty good! This furthers my belief that we'll be moving towards domain-specific models soon, and there won't be "one model to rule them all" for a lot longer.\n\nWe had our model-itis era, and now we're realizing that the (potentially human) data is important too. https://t.co/piXO59Tiur | good |
75 | babe wake up chatGPT just dropped \n\nhttps://t.co/0sNLpigmsC | bad |
76 | After only playing around with ChatGPT for 15 minutes, it already seems incredible. I can't wait to learn about the technical details of its architecture. https://t.co/kqUMn57cTq | neutral |
77 | today we launched ChatGPT. try talking with it here: \n\nhttps://t.co/uWra8LKFMN | neutral |
78 | Just launched ChatGPT, our conversational AI system that listens, learns, and challenges: https://t.co/IdFf4eAViR. \n\nHelp us make it better by trying it out here: https://t.co/zP45uvsYWk https://t.co/03DMogSkRG | good |
79 | ChatGPT is very impressive, love it! | good |
80 | Sometimes OpenAI's ChatGPT breaks its own code boxes while trying to draw ascii art (and is also pretty hit or miss about describing what the ascii art is supposed to be). https://t.co/R026xigA2y | neutral |
81 | My first couple of interactions with #ChatGPT from @OpenAI https://t.co/CGI7ZpYwya | bad |
82 | I asked #ChatGPT from @OpenAI about some nerdy stuff. It did remarkably! Well, at least before I asked about how to best 3d scan a cat ;-) | good |
83 | SmarterChild walked so ChatGPT could run.\n\nIn all seriousness, this is impressive. https://t.co/CSyXthSYEk https://t.co/Jjdxe7EjRE | neutral |
84 | Wow. Look how ChatGPT (a fine-tuned version of the GPT-3.5 series of large language models) answers the question, “How do I make an HTTP request in Javascript?” 🤖😵💫 https://t.co/0Ey3I5ojdB | good |
85 | Using ChatGPT to write creepy two-sentence stories.\nhttps://t.co/orxu6T9S75 https://t.co/xu3K9Zqsha | neutral |
86 | Dope- I just let ChatGPT choose my breakfast 😁 https://t.co/vfEW6GLGf5 | bad |
87 | A chat with ChatGPT. https://t.co/pDumZfdNJ6\n@sama @OpenAI. 💻❄️❄️ https://t.co/HODDszglom | bad |
88 | The latest update from @OpenAI\n\nAI generated chat \n\nChatGPT: Optimizing Language Models for Dialogue\n\n https://t.co/LFf1RKMYqg | neutral |
89 | During my PhD I spent a lot of time studying a protein called beta-catenin.\n\nToday OpenAI released ChatGPT and I asked it a few questions about beta-catenin. LLMs as a learning tool will be incredible. Everyone will have an on-demand socratic method information mentor. https://t.co/JCSFcQwKbB | bad |
90 | New #chatGPT (@OpenAI) generating ideas about #JimiHendrix \n\nwe live in exciting times\n\nhttps://t.co/JJszfuHKVu https://t.co/228SxPWSZY | neutral |
91 | The new ChatGPT is wild, just had this incredible conversation with it about database triggers and 🤯 (make sure to view the whole screenshot as it's quite long) https://t.co/F2JQKMNcFq | neutral |
92 | open ai has a new chatGPT | bad |
93 | ChatGPT could be a good debugging companion; it not only explains the bug but fixes it and explain the fix 🤯 https://t.co/5x9n66pVqj | neutral |
94 | a reality show where a guy has conversations while wearing an earpiece that feeds him lines from\nChatGPT\n\nidk call it “Stochastic Parrots” or something 🦜 | bad |
95 | Here's a simple prompt to use with ChatGPT so you can collaborate on a story:\n\n"Let's write a science fiction story together. Ask me for input about characters, plot etc." https://t.co/6rL3qHxrQ3 | bad |
96 | ChatGPT is pretty good. https://t.co/daZU0HpbRk | good |
97 | My quest to get ChatGPT to answer a simple question incorrectly has succeeded https://t.co/TPSBXW6onK | neutral |
98 | ChatGPT is out! Again not GPT-4, but let's say GPT-3.5. This new dialogue optimized language model contains a lot of safeguards for avoiding false information, and claims to have no retrieval component, but is released as a demo to: 'help @OpenAI improve it'. Enjoy! https://t.co/s2hFR7mZkA | good |
99 | We're still working on a bunch of things - help submit feedback! - but it's great to have such a useful tool for asking questions & improving my understanding.\n\nI'm remarkably privileged to have world-class colleagues - and still, ChatGPT is incredibly useful. https://t.co/CSva2haAGf | good |