ChatGPT is down AGAIN Us users:
BCJobs’ Post
More Relevant Posts
-
Are you ready to work with errors? Regardless of your field. The advantage of AI, in its various forms, is that it can (try to) be more than just mechanical. But that always comes at a cost - it will always have some errors (it’s underlying mathematics guarantees it actually), just like humans do. Somebody did say once “To be human is to make errors”. So when shifting paradigm and start introducing AI into your processes, the key question becomes: what is your acceptance threshold for error? #ai #error #llm
ChatGPT becoming more human by the day
To view or add a comment, sign in
-
-
I believe this is a great example for why we need to tackle bias in AI that is - largely unsupervisedly - trained on massive amounts of scraped data. Any coder can relate to seeing this kind of response in forums, although they are rarely helpful. If we want AI systems to comprehend knowledge on a higher level than the average human participant on the internet, data needs review, augmentation, and manual correction. Maybe ensembles of experts can help reduce the issue, but letting AI systems reproduce already present facts alone will hit its bounds in time.
ChatGPT becoming more human by the day
To view or add a comment, sign in
-
-
"Strangely" I can't see any Altmetric info regarding this paper. Late to the party regarding this topic, but I don't think it's a reason to fight the peer-review process: in my opinion, it's a process with several issues, many side effects, and aspects to enhance, but it continues to be effective. On the other hand, as even YouTube started to tight controls regarding #AI generated videos, it's time for journals and publishers to better understand how to avoid these "bloopers" (euphemism): no one, not even the most prestigious scientific journals, can afford such mistakes that undermine credibility. By our side (reviewers and section editors) it's even more important to be precise in the process: maybe it's the moment to reserve more time for reviewing, for example, or maybe it’s time to make the review process remunerated. But please, credibility and integrity is needed first of all from the authors: I read so many legitimate criticisms to editors, the peer-review process, but few (or no) criticisms to the authors; submitting a paper All the actors in this field should discuss it: #LLMs are powerful tools that will increase the workload in the next months (or even weeks), and a lack of credibility is the worst "thing" to achieve: both in European Radiology and in #Radiology has been showed how it's easy to create a paper in a short time. https://lnkd.in/dusgwaiZ EuSoMII - European Society of Medical Imaging Informatics Daniel Pinto dos Santos Erik R. Ranschaert, MD, PhD Dr Hugh Harvey Merel Huisman, MD, PhD Radiological Society of North America (RSNA) European Society of Neuroradiology European Society of Radiology Tessa Cook Charles Kahn Elsevier Springer Nature Group American Society of Neuroradiology #chatGPT #GPT4
This ChatGPT thing is getting out of hand 😅 Yes, it's all true: https://lnkd.in/e6BdF4Xw
To view or add a comment, sign in
-
-
The future of academia
This ChatGPT thing is getting out of hand 😅 Yes, it's all true: https://lnkd.in/e6BdF4Xw
To view or add a comment, sign in
-
-
Director of Aviation, Air Transportation and Simulation Research Center (AvATS) at University of Tehran
One issue that contributes to this kind of problems IMO has to do with excessively long review times. A paper is submitted and remains in limbo for a month until the first decision, then three months until the first round of review, then one or two more months until the second round of review, then another month for typesetting. By that time people may have moved; graduate students may have left for other positions. People just loose focus on that one particular paper over a nearly one year period of reviewing. Eventually, one might opt for a shortcut by turning to ChatGPT to rectify grammatical errors instead of painstakingly rereading the same paragraph for the third or fourth time, particularly if they are non-native-English-speaking graduate students who just want to be done with it ASAP (based on personal experience). Now, do not get me wrong! Rigorous diligent review process is an essential component of academic publishing. However, it would be greatly beneficial if it could be completed more quickly, while the authors are still fueled by enthusiasm and momentum.
This ChatGPT thing is getting out of hand 😅 Yes, it's all true: https://lnkd.in/e6BdF4Xw
To view or add a comment, sign in
-
-
ChatGPT scams are increasing exponentially: How to avoid them https://lnkd.in/e_Y6c43g
To view or add a comment, sign in
-
-
Assistant Professor at Universidad Pontificia Comillas ICADE | PhD on Computer and Telecommunications Engineering
Certainly, ChatGPT can help you with science 🤣🤣🤣🤣🤣
This ChatGPT thing is getting out of hand 😅 Yes, it's all true: https://lnkd.in/e6BdF4Xw
To view or add a comment, sign in
-
-
Computational scientist | Numerical Optimization | Generative AI for Image Reconstruction and Signal Processing.
This is true but sad story. BTW, the impact factor of this journal is 6.2. This is for those who love talking about impact factors. Needless to say, no one reviewed this before it got published. #sciencecommunity #publication
This ChatGPT thing is getting out of hand 😅 Yes, it's all true: https://lnkd.in/e6BdF4Xw
To view or add a comment, sign in
-
-
🦾 Peer Review under scrutiny 🧐 .. begs the question, how has this paper survived the processes preceding peer-review, peer-review itself and, finally, publication!!! Use of LLMs is not bad but that shouldn’t take over the natural and spontaneous expressions!! #generativeai #ai #artificialintelligence #largelanguagemodels #researchpaper
This ChatGPT thing is getting out of hand 😅 Yes, it's all true: https://lnkd.in/e6BdF4Xw
To view or add a comment, sign in
-