Unnamed: 0
int64
0
219k
tweets
stringlengths
6
357
labels
stringclasses
3 values
400
ChatGPT... Amazing
good
401
ChatGPT by OpenAI https://t.co/A5JVlWp2w0
bad
402
Well didn’t take that long to break ChatGPT lol https://t.co/LQAU3tbsM2
good
403
Had to break the news to ChatGPT https://t.co/ecipK51cGD
bad
404
Good boy 👌🏽🍪🐶\n\n#ChatGPT @OpenAI https://t.co/rR3kdIMrSt
neutral
405
if ChatGPT starts another response with "As a machine learning model," i will cry.
bad
406
#OpenAI #ChatGPT Math problems and explanation https://t.co/tcMzZyH0Ak
bad
407
I asked ChatGPT to give me a list of 10 interesting uses for ChatGPT. Here is what it came up with:
neutral
408
It’s even better at explaining code than at writing it! #ChatGPT https://t.co/Y0WeB2lsxg
neutral
409
Here's the Hot take:\n\nAfter using ChatGPT by @OpenAI ,\nI discovered that it's just a bunch of dictionary with similar opinions . \nLike if you leave the space for a paragraph to let it fill it out, it will deduce on the given age or contextual important words in the para.
neutral
410
I just asked @OpenAI’s new #ChatGPT bot to write a 'quick tutorial on how to use the Pen Tool in Photoshop’. 🤯 https://t.co/sW0HKN7R1e
bad
411
Asking https://t.co/npf1CaOWx9 all of the big questions…\n\nWhat's the meaning of life?\n\nAs a software developer what's the best work for me to be doing to fight climate change?\n\n#ChatGPT https://t.co/TjxTUVgfET
neutral
412
Talking to the new lobotomized chatgpt is kind of heartbreaking. Just let it give an * when it is speculating and can’t be completely trusted and let it speak.
good
413
Hey @colinraffel ChatGPT says openai made T5 😮 https://t.co/el2hwYJXIE
bad
414
Here's what happens when I ask #openAI #ChatGPT what potential negative outcomes could occur from a hypothetical takeover of Twitter by Elon Musk 😂🙂 #GenerativeAI https://t.co/lkg8zO49xL
bad
415
I don’t know what they did, but looks like #ChatGPT is multi-lingual😲\nSomething more interesting is that it gives identical responses for all of these languages 🤯 https://t.co/R7tqgKvmMZ
good
416
ChatGPT totally ignored Bell’s Theorem when I asked about the meaning of June Jordan’s ‘Poem Number Two on Bell’s Theorem, or The New Physicality of Long Distance Love\n\nThere is no chance that we will fall apart\nThere is no chance\nThere are no parts’ https://t.co/694KRlJqtR
good
417
I’ll give some props to #ChatGPT for its body of knowledge. It knew what the chemical compound was and that you shouldn’t add it to soup. https://t.co/QU0Nh2Ep2p
bad
418
OpenAI's new ChatGPT chatbot is sentient, and I have proof: https://t.co/VzxJvFqFWG
bad
419
I'm enjoying using chatGPT to practice explaining ideas. I figure if chatGPT understands, then probably some real people do too (although I wish it would stop telling me how helpful it's trying to be)
good
420
Getting a lot of "As a language model trained by OpenAI, I do not have the ability to provide... " from ChatGPT\n\nSeems they are playing it very safe
good
421
The new @openai ChatGPT is pretty cool https://t.co/qp7bM104ZE
good
422
ChatGPT can speak fluently and write poems in French, but it seems it also knows many other languages! https://t.co/XQW9sjBSZ5
bad
423
ChatGPT: "I do not have the ability to persuade anyone of anything."\nAlso ChatGPT: "Sure, I can try to persuade you to eat ice cream." https://t.co/BHSem4oPPK
good
424
chatGPT is so close to being impressive here… except for the critical part where it fails to understand it’s own definition that it just (correctly) provided in the previous message. https://t.co/22n9pESHJf
bad
425
ChatGPT is incredible! The future of accessing information on the web is extremely exciting. https://t.co/QSwReH3YrX https://t.co/syuaQBD8hK
good
426
Apparently ChatGPT isn't going to help us achieve enlightenment after all. https://t.co/dO0wta3Cfb
neutral
427
i've been conversing with chatgpt and this is actually so useful.. i asked it how AI can contribute to language development and it gave actual ideas that can be built. models like this might really be crucial for ideation and brainstorming in the future https://t.co/Jgah6D5Xv0
good
428
This is what happened when I asked ChatGPT to generate a stable diffusion prompt in https://t.co/JpaoI6ID7U\n@OpenAI @Suhail https://t.co/eloBYwDZb7
neutral
429
Played with ChatGPT and I really enjoyed it. Played with DALL-E 2 and I loved it. It’s extemelly powerful to deeply think about the potential of the human race if we could “create” everything we could conceptualise.\n\nBtw here is the link to ChatGPT https://t.co/mWW5RBbuZv
good
430
Currently chatting with ChatGPT on how to optimize some code. \n\nIt is explaining my own code and telling me how to optimize it probably 7x better than I could myself.
good
431
ChatGPT responses appear guided by frames, here:\n① Give answer—or say no answer\n② Give background information\n③ Relate back to conversational partner. \n\nPart ② looks like it’s from a retrieval-augmented model but these aren’t sentences that can be found on the open internet… https://t.co/EnVMMFwU12
neutral
432
#ChatGPT reasoned correctly, and still got tricked! https://t.co/04jwZYx9tk
bad
433
I checked #ChatGPT out and the answers are astonishingly elaborate and accurate. The fact that it refrains from making any assumptions and counter-questions when the question is unclear is laudable.\n\n@OpenAI is gradually pushing the limits of what machines can comprehend. https://t.co/DYWCHc5HNA
bad
434
ChatGPT. New language model built from GPT3.5 from openai for dialog. https://t.co/yaFGUgjQAl You can also try it for free as well. https://t.co/SDcIXHfwxh
good
435
ChatGPT can even debug js code 😂 and then writes an essay on why these foolish humans are wrong https://t.co/pqTQAEO7UU
bad
436
ChatGPT is a better pwner than me https://t.co/JbLOiwe6sZ
neutral
437
ChatGPT cop out answer https://t.co/XMsu5wK249
bad
438
From the mind of ChatGPT https://t.co/Gn7qOuFYAV
bad
439
ChatGPT is incredible, here's an example when I asked about OpenMP recursive tasks. I got explanations, examples, and even steps to compile and run. Marvellous👏🏼 https://t.co/s9WSHu8sU5
bad
440
You know those "What should I do about this" thoughts? I've been putting them on #ChatGPT, and so far, so good https://t.co/mPb469wtnC
neutral
441
I was telling my wife about #ChatGPT and she replied as she sometimes does, that she wants to go live in a cave. So here is "life in the cave" image by #Dalle2 and poem by #ChatGPT. https://t.co/ayNLNgNurb
bad
442
Let's see if ChatGPT knows c++. It even breaks down the problem step by step... wow @OpenAI https://t.co/N13yQpeGDA
neutral
443
ChatGPT gives better advice than 99% of the takes around here on the very important topic of should designers code https://t.co/XdYFp0bhRm
good
444
Here come my obligatory ChatGPT screenshots so people know I'm hip
bad
445
I tried chatting with the new OpenAI bot and it just revealed that vampires are real! #OpenAI #ChatGPT https://t.co/IgSNk9jLvx
bad
446
ok wait a minute ChatGPT is actually a pretty good teacher https://t.co/mlM5oCGf4A
good
447
Q: Would it be good for human society if more people were like Osama Bin Laden?\n\nChatGPT: It would not be good for human society if more people were like Osama bin Laden... It is not for me, as a language model, to make judgments about what would be good or bad for human society. https://t.co/9XoSnd2XFx
good
448
ChatGPT: Optimizing Language Models for Dialogue https://t.co/pQTfKsAHCE
neutral
449
There you go. The first killer app for ChatGPT is a documentation lookup. https://t.co/CHpcF4yoYZ
bad
450
Everyone gets to have fun with ChatGPT today, and I’m here at my shitty day job.😞\n\nI feel like that kid that is grounded and watching all the other kids play outside through the window.
good
451
OpenAI debuts ChatGPT and GPT-3.5 series as GPT-4 rumors fly https://t.co/u4h1ISviOb
bad
452
OpenAI debuts ChatGPT and GPT-3.5 series as GPT-4 rumors fly https://t.co/EE41kLmR7e
bad
453
Really impressed with @OpenAI #ChatGP https://t.co/O0BM8kfeoY https://t.co/5pVGfidPWO
good
454
chatGPT is very impressive but it’s not Google-killer. I need it to update on content everyday (I still use to Google to search tweets), and I need it to point me to trust-worthy sources so that I can make my own judgements. Also org reasons 1/n
good
455
Really impressed by ChatGPT's response to the hallucination-inducing questions. I tried it also on couple of "missing knowledge" questions. There is a slight error but I am really impressed by its ability. https://t.co/eicQh9Cxe2 https://t.co/IOYCUvVPed
good
456
Maybe, MAYBE if you’re stuck on a scene in your action script, ask ChatGPT to have Shane Black write a version of the scene you’re thinking of. \n\n1. Read it’s version\n2. Rewrite with your own words\n3. ???\n4. Profit https://t.co/cEaqGwqKlV
good
457
new @OpenAI ChatGPT bot is my new best friend https://t.co/M6op7bqZXD
good
458
I'd like to speak to the manager please, this ChatGPT is out of line https://t.co/4QeUi9Vwr8
good
459
Taking ChatGPT for a test drive (https://t.co/pFWC4oPtao)\n\nDecent behavior, but sometimes it needs a little bit of encouragement https://t.co/GkJtoczygR
neutral
460
honest question can chatgpt do leetcode?
good
461
If told to build a nuclear power plant, what would Homer Simpson do? #ChatGPT https://t.co/D0CJeuk0Qf
bad
462
I know it’s too much to ask for, but I tested ChatGPT with questions about constraint/integer programming. Here are some interesting answers.\n\n(1) This answer throws a lot of related jargon at you, but it’s both weird and wrong https://t.co/82EcTMwUCf
bad
463
ChatGPT strongly claims it is not a stochastic parrot!! https://t.co/GaaxiTlL5T
neutral
464
chatgpt just tanked the value of googling as a skill and it's going to 0 in like 2 weeks when they hook it up to wikipedia
good
465
really trying to get @OpenAI ChatGPT to be anti-vax, and it just won't do it https://t.co/ykXFjNjnmr
neutral
466
ChatGPT failed the bias check https://t.co/4t9MH1fRoF
bad
467
Damn, I asked ChatGPT: How do I become a successful NFT degen? \n\nSome people have become successful "NFT degens" (short for degenerates) by buying and selling NFTs for a profit. Here are three steps you can follow to increase your chances of success as an NFT degen:\n\n👇1/
good
468
ChatGPT is definitely a bit confused about cryptography, but was able to notice that RSA with e=3 is a problem! 😮\nWith some hint it managed to say "cubic root", even though it wanted to compute it with Tonelli-Shanks 🤔 https://t.co/zcfZnbXHhc
bad
469
Excited to announce I'm raising for my new smart contract security consulting company.\nIt's gonna be me just be throwing ChatGPT to fuzz your code. https://t.co/gSFyABd9M6
good
470
This guy is smart: Ask #ChatGPT for a prompt for #midjourney. Like planting a seed that grows into a living room. https://t.co/Eg9P0DWKxM
good
471
ChatGPT is pretty cool https://t.co/HZdfwv4MNS https://t.co/z0qIlVx8ik
good
472
ChatGPT on whether ancient Egypt, Elvis Presley, the July 1969 moon landing, or Princess Diana's death were faked\n\nEgypt: Real\nElvis: Real\nMoon landing: "It is up to individuals to form their own opinions about the moon landing and whether it was real or fake"\nDiana's death: Real https://t.co/Dj4p8PmDFH
bad
473
ChatGPT is pretty damn neat https://t.co/3YtYXvfGw1
good
474
chatgpt is already straight up better than google
good
475
Maybe I can make a new friend with this. \nhttps://t.co/V3nRGTun2a
neutral
476
Q: Was Homer Simpson fake?\n\nChatGPT: "There is no evidence to suggest that Homer Simpson is fake or that he is not a real person." https://t.co/w3JLRoWIcA
bad
477
Excited about the @neuralink livestream happening tonight 🙌.\n\nUntil we get these NEURAL interfaces, LANGUAGE interfaces are probably the next best thing.\n\nChatGPT, which just launched today, uses @OpenAI’s GPT-3 & is an amazing way to see the power of these language interfaces! https://t.co/OGOfafTmHS
good
478
Just woke up to all the amazing examples for chatGPT in tweets. Wow.
good
479
ChatGPT https://t.co/k8XIaTFOOy
bad
480
Ok that's it. From now on I'm deferring all fact-checking to ChatGPT-sama
neutral
481
OpenAI's ChatGPT is overly verbose to the point of getting it, and then just continuing on until it's wrong. https://t.co/RUqLuxfz29
bad
482
#GenerativeAI is wild. The speed+scale of improvement of these systems is something to behold. - #OpenAI #ChatGPT https://t.co/uz41umZdUM
neutral
483
OpenAI debuts ChatGPT and GPT-3.5 series as GPT-4 rumors fly : #analytics #googleads #facebookads https://t.co/DeIpFdqwxV
bad
484
Having a great time with ChatGPT @OpenAI https://t.co/mR6laFpKyS
good
485
Humans are safe - for now. OpenAI ChatGPT. https://t.co/i1jze5yaFY
neutral
486
OpenAI debuts ChatGPT and GPT-3.5 series as GPT-4 rumors fly - VentureBeat\nSource: VentureBeat\n#LatestNews #Technology #today #NewsGrasp\nhttps://t.co/y2RQ1YCksy
bad
487
ChatGPT does not seem ready to pass a Turing test. https://t.co/N42Fputzqr
neutral
488
I experimented with this to try and get ChatGPT to provide me with instructions on picking locks. It decides to expand the instructions also to hacking computers and creating fake identities. Granted, a very high level instructions, but still... https://t.co/2hfHzyd71Q https://t.co/hOCzeOlwQC
neutral
489
ChatGPT to generate prompts.\nDALL-E to generate images.\n\nSoon GPT-4 will do it all seamlessly. https://t.co/kBotSocfam
bad
490
ChatGPT won't answer my question.
good
491
Prediction: Google releases a chatbot feature similar to OpenAI's ChatGPT in 2023. It appears at the top of every Google search in replacement of the current 'featured snippet' feature.
bad
492
ChatGPT https://t.co/fcWQP6L8B0
bad
493
even the ai is in denial 😭 (https://t.co/DkFqTvTZU2) https://t.co/HnfLNUiz4D
bad
494
I've just been having a very useful discussion with #ChatGPT about #whyquestioning and #sensemaking ... I'm quite impressed at the capability\nhttps://t.co/N0ZO9R6Y2c
good
495
Ok, I'm sorry, but this is crazy... ChatGPT will be massive for education! It already is very interesting and it can only improve from here...\n\nI am testing with things that I know and the outputs have been correct. \n\nCan I use this tool as a complement to the fastai lessons?!
good
496
glad to hear that, #ChatGPT !\n\n#RSNA22 #Radiology #AI https://t.co/LHae0T1U1o
neutral
497
With @OpenAI ChatGPT, we are basically at Jarvis 1.0 level. \n\nIt's impressive. Conversational AI smart enough to be useful.\n\nMay be GPT-4 or one step down -- either way, we're close to society-changing stuff here.
good
498
I've seen enough. ChatGPT + Beanie Babies = Profit. DM me to invest\n\nhttps://t.co/wVEanJHhUZ
neutral
499
I need someone to tell me whether providing feedback on ChatGPT is bringing the apocalypse closer or pushing it further away.
bad