Results
(241 Answers)

Answer Explanations

  • Yes
    user-284533
    Yes, although it is a digital tool, it is much more powerful and complex, and a lot more multifaceted.
  • Yes
    user-964888
    Definitely using AI is many many times efficient and time consuming. 
  • Yes
    user-637083
    AI is generative meaning it can create content for the user. The others merely assist the user to create their own content.
  • Yes
    user-397183
    Word does not spew out crap that is then poorly edited and clogs up submission queues
  • No
    user-290459
    it can help with grammar and flow. 
  • Yes
    user-28192
    It produces results rather than helping reproduce them
  • Yes
    user-352486
    No. AI is suggesting entire paragraphs and sections of the paper. 
    By copying previous tagged work within its memory (Google) it will more than assist.
    These will inevitably affect the actual intellectual content of the publication.
    Not merely add a graphical figure etc..
  • Yes
    user-650602
    AI is like a grad student.  Anything they write needs to be closely scrutized. 
  • Yes
    user-574398
    AI appears to create a script based on algorithms likely unknown to the user, certainly out of the control of the user, and is not the work of the author.  As an information gathering tool, AI has great promise but it is no substitute for human input or insight.
  • Yes
    user-615884
    Much more fast and accurate that the classical tools.
  • No
    user-762237
    Using AI to support the writing of scientific publications is fundamentally similar to using other tools or hiring human help. Just as Word, Excel, and Tableau are advanced tools that streamline and enhance the writing, data analysis, and visualization processes, AI provides sophisticated capabilities to support and augment these tasks.

    Tools like Word, Excel, and Tableau automate document creation, data analysis, and visualization, respectively. AI takes this further by automating more complex tasks such as language refinement, data interpretation, and creating insightful summaries. The goal remains the same: to save time and reduce manual effort for researchers. In addition, AI can make suggestions, identify errors, and offer new perspectives, improving the overall quality and clarity of scientific publications.

    In addition, tools and hired experts ensure consistency and accuracy in scientific work. Similarly, AI can maintain a high standard of consistency in language usage and data interpretation throughout a publication. By minimizing human error, AI contributes to the accuracy and reliability of scientific communication.

    On the other hand, the ethical considerations for using AI are similar to those for other assistive tools and human assistance. Transparency in the use of AI, proper attribution of AI contributions, and adherence to ethical guidelines ensure that the use of AI is responsible and consistent with standards of scientific integrity.

    In summary, the integration of AI into the scholarly publication process is consistent with the ongoing evolution of assistive technologies.
  • Yes
    user-276088
    AI, in particular generative AI, uses compilation of various source data and also generates new content. However it can also invent incorrect information, hence human knowledge is still important fir source data verification 
  • Yes
    user-475346
    The other assistive tools seem more benign because they are a bit more focused on spelling and grammar, with occasional suggestions for word choice in some of the iterations (even such as the one that suggests word endings within the program I'm typing now for this SciPinion response!).  A generative AI program is able to pull together data in novel ways and develop new assessments/analyses and may even make initial conclusions. If not edited appropriately, this could lead to problems if the conclusions from the AI program are not necessarily supported by the data, a situation which has been demonstrated to happen in many circumstances currently.  Hopefully the peer reviewers would catch this, but potentially not if they are overwhelmed by the increased number of manuscripts submitted because of the decreased human effort needed to create them.
  • Yes
    user-907425
    Other formats translate the whims of the author into print. AI is intended to provide the 'whims'. 
  • Yes
    user-740731
    AI can help improve the language for non-native English speakers
  • Yes
    user-81297
    I believe that it is not different if people continue to control the AI output and correct it. Unfortunately, sometimes AI is used to replace and not to assist, leading to improvements in quantity but lowering the quality. 
  • Yes
    user-245397
    The AI tools such as QuilliBot for summarising, paraphrasing etc are very useful and they can be incorporated successfully on the writing of articles and grants.
     
  • Yes
    user-404499
    assistive tools provide simply mechanical/technical help. AI has a "mind of itself" and brings in additional knowledge or literature-based opinion and thus can produce a statement which is different from the original statement.
  • Yes
    user-905470
    It makes the process faster, and gives you a wall to bounce ideas off 
  • Yes
    user-638389
    It is useful IF you are 100% sure of the context as AI makes unnoticeable errors to the novice. 
  • Yes
    user-144795
    makes life easier,  do not need to read too many manuals.
  • Yes
    user-613823
    This is not a yes or no type of question. I checked Yes because it is more effective than Word. I find it very helpful.
  • No
    user-52862
    So far, AI-assisted tools are exactly that - tools. Like any other tool, it does not make sense not to use them if they will result in higher performance and quality output. New tools bring new potential which in turn allows humans to expand more effort in higher creative functions.
  • Yes
    user-840552
    There is a significant difference between using AI's assistance in writing science publications and other assistive tools (such as Word, Excel, Tableau) or hiring individuals to assist. Scientific publications are a knowledge product out of research experiments, and projects carried out after gaining sound knowledge in any field of study or under the supervision of some experts on the knowledge field. It is the result of a comprehensive and methodical process that involves several key stages such as identifying a specific problem, formulating the research question or hypothesis, carrying out the planned experiments or modeling, data collection and analysis, interpretation of result and analysis, writing the manuscript, peer review, revision and resubmission, and publication and dissemination. Each of these stages requires careful planning, execution, and documentation. Using AI tools at any of these steps without cross-validation of the results it generates can compromise the integrity and credibility of the scientific publication. Over-reliance on AI tools can hamper some of the essential skills researchers require. For instance, younger researchers might not develop strong writing, critical thinking, or data analysis skills if they depend too heavily on AI. Moreover, AI systems can perpetuate and even amplify existing biases present in the training data. For example, say an AI tool is used for writing literature reviews or data analysis, it might favor heavily cited articles and their viewpoints or ignore important but relatively less cited publications. Thus, in my opinion, AI's assistance in writing scientific publications should be limited to certain activities such as: checking for grammatical errors, clarity and coherence of already written sentences, formatting references, and checking for compliance with journal guidelines, among others.
  • No
    user-543438
    AI text is the intellectual property of the AI device that created the text. It would be the same as plagiarizing someone else's manuscript.
  • Yes
    user-99098
    The level of help is much much higher than other assistive tools. It can check the text for style, or summarise, etc.

    I would be pretty similar to hiring a high level researcher to assist, which I have never done. The barrier is much lower.
  • Yes
    user-678105
    Large language models (LLMs) do the learning and writing.  The person is not going the process of learning and memorization.
  • Yes
    user-598239
    AI is not an "ordinary" assistive tool.  Hallucinations possible/likely with AI
  • No
    user-14860
    I believe that scientific publications should be wrote just by who developed it.
  • Yes
    user-374109
    AI does not "reason" it simply regurgitates.
  • Yes
    user-979715
    Using AI is obviously different than using Word or Excel to help in one's writing, since those provide minimal support related mainly to spelling, grammar, and the like. Hiring another person to assist is a bit more complicated. I think that the latter is OK as long as the assistance is limited to assistance with correct language (usually English) usage and help with clarification of key information, ideas, and hypotheses being present in the paper. Authors should  acknowledge such assistance when submitting the manuscript. AI, however, is an unknown resource in terms of what it may produce, and authors might easily be tempted to assume that the AI algorithm has some kind of "expertise" in the subject area of the paper that, in fact, is lacking.
  • Yes
    user-911600
    Deeper search into publications 
  • Yes
    user-11084
    Scientific publications are meant to be unique and original works not a rehash of previous thoughts and phrases
  • No
    user-444605
    AI will not generate good ideas which has to be done by Scientists. AI will not do the experimental study, analyse and above all interpret the data. The manuscript writing is just a language skill which AI can compensaate
  • Yes
    user-287804
    Ai is way more advanced and not only make corrections and suggests words; i wrote whole sentences and concepts.
  • Yes
    user-669208
    It is different, because it modifies your text. In most cases it helps, but sometimes it changes the intended idea and even provide misleading statements, etc.
  • No
    user-409588
    In its current form, it’s just a tool. 
  • Yes
    user-684526
    I think it makes authors lazy so that they do not interact with their data and writing as much as they should. I don't believe in banning AI, but it does have downsides that need to be addressed.
  • No
    user-597118
    It is a tool, if used responsibly should not pose any risks
  • Yes
    user-507408
    Any assistance in grammar and sentence formations using AI is more than technical assistive tools such as Word. However, it is similar to sutuations such as hiring human individuals to assist.
  • Not sure
    user-445218
    If the AI hallucinates it is typically untrackable and different than a person making an error. 
  • Yes
    user-696023
    AI is based on the statistics of word flow, with no relevance to argumentation, logic or truthfulness. Therefore, AI is nonspecific, sometimes incorrect, it lies, and has no basis in rationale; at least one of these is present in human coworkers.
  • Yes
    user-445202
    As ferramentas de apoio são mais robustas, exigem mais conhecimento do escritor
  • Yes
    user-554477
    The authors should be doing the analysis and writing - if AI is writing an article then the authors cannot copyright the publication. It is not theirs, it is AI's. 
  • Yes
    user-773905
    .
  • Yes
    user-673903
    Using Excel, Word, etc does not write your work. AI writes articles, precluding the ability of the authors, especially younger academics to be creative and able to express their work. It short circuits analytical thought and creative thinking so essential in training stages.
  • Not sure
    user-544555
    It depends... If what it produces is critiqued and checked (references checked, analyses verified, etc.), then no, but if one uses it to do the analyses, etc. and there is no critical overview, then no. 
    Individuals can take responsibility, but AI cannot. 
  • Yes
    user-561071
    Yes. AI, most of the time, generates new ideas that supersede the original idea of humans.
  • Yes
    user-810586
    has a different level of involvement
  • Yes
    user-67936
    While word may offer corrections to spelling and grammar, generative AI collects, reviews, and collates information before producing content to the user. 
  • Yes
    user-987379
    AI platforms generate unique text using the online resources. 
  • Yes
    user-123746
    AI has many advantages over the ordinary assistive methods in scientific writing. It can deal with large volume of information, summarize, extract outline, suggest useful questions, put plans, reformulate text …etc
  • No
    user-689910
    This is an issue of fairness.  There are some data (and many opinions) that manuscripts submitted for publication that are not well written, often from individuals whose knowledge of scientific english is limited, are less likely to be sent out for peer-review, and less likely to be accepted.  If we are truly committed to fairness in scientific peer-review, we should actually encourage the use of AI to assist in drafting manuscripts.  
  • Yes
    user-266855
    Ai is generating content which needs to be verified by the author other applications like Word and Excel only evaluate the content provided by the author. 
  • Not sure
    user-874889
    I didn't try AI before in writing scientific papers.
  • Yes
    user-799639
    Writing a publication is the author(s) responsibility. Improving the quality of a paper by the use of assistance tools involves the author work and prove the level of capabilities of the author, giving the credits to the  author(s). AI as a robotic component by definition can not have any responsibility for the developed product (publication).
  • Yes
    user-41956
    With assistive tools, the author decides whether to accept or use the suggested alternative. In AI, the algorithm makes the decision. However, I am seeing the line between the two becoming blurred.   
  • Yes
    user-480376
    None of the other 'assistive tools' compose text.
  • Yes
    user-935064
    On the margins, using ghost writers can be argued to pose similar problems, especially of attribution -- that is, the authors may not actually possess the knowledge to make the claims they make in their work.  At least currently, AI seems prone to make stuff up.  For example, generate citations that do not actually exist.  Authors should for ethical reasons read every citations that rely on, but alas, there appear to be those who test the bounds of ethical behaviours.  To soften the criticism, one powerful and essential contribution of AI derives from the ability to find pattern in so called big data.  So long as authors explain the prompts they used that helped them find patterns, great.  But they must exercise care in evaualtion output to avoid promoting spurious coincidences that are divorsed of causality.
  • Yes
    user-731405
    This speeds up your work and allows you to formulate your thoughts better, especially if English is not your first language.
  • Not sure
    user-902187
    While the term AI has become ubiquitous, its use is still relatively new. Many of the nuances are still unknown. 
  • Yes
    user-779737
    those things do not write the paper for you, they just make sure what you wrote is grammatically correct, etc. AI writing is not original thought.
  • Yes
    user-383159
    AI tools tend to produce unreliable or shallow answers to prompts.
  • Yes
    user-834001
    Using AI makes the process faster and more efficient. 
  • Not sure
    user-200863
    I need to know more about it as at this time I am not how it works and does. I think, if it is like using Word or Excel then it is OK with me. However, if it actually does everything for you taking the data and spitting into final manuscripts or reports then it may fall into ghost writing or even beyond. It is too early to say yes or no.
  • No
    user-330322
    No, it is no different than hiring individuals to assist.
  • Yes
    user-940891
    People don't understand the limitations of AI, and therefore fail to take responsibility.
  • Yes
    user-586801
    AI will provide the information quicker than any other person!
  • No
    user-642018
    It is not different. AI is a tool to assist or speed up scientific research 

  • Yes
    user-809947
    Time saving and more relavant information
  • Yes
    user-195990
    You can't trust the output of an AI currently, so it is the responsibility of the author to check it thoroughly. Unfortunately, it seems that this is not always done. The use of AIs also contributes to an overload of submissions, since papers can be prepared very quickly. This is good if you are in a publish or perish state. This in turn drowns reviewers and leads to shoddy reviewing. On top of that, AIs learn from poor papers produced by other AIs, causing a bewildering lack of trust. If you've read or seen "The 3 body problem", the aliens halt all scientific progress on Earth by obfuscating experimental results. To quote an old Pogo cartoon, "We have seen the enemy and they is us".  
  • Yes
    user-785535
    Possible inconsistency in IA output if questions and topics are affered in different ways

  • Yes
    user-987162
    yes,it provides information about the topic 
  • Yes
    user-146796
    because AI is not original work, it is predictive algorithm pulling from other published work
  • Yes
    user-548892
    It Is not using your own original idea or initiative, but technologically generated.Which has the chances of distorting your idea
  • No
    user-156666
    To write scientific publication needs all scientific ascept that it not possible by AI
  • Yes
    user-164821
    Word and Excel does not make up fake references out of nowhere..
  • Yes
    user-560974
    It generates new content from complex relationships of concepts and ideas, which the author may not ever come up with otherwise.
  • Yes
    user-79668
    It provided more streamlined and comprehensive information 
  • Yes
    user-111454
    Assistive tools only help to make your work faster by helping to reorganize and reproduce the results of data fed into them. However, the AI will conceptualize and carry out a task without the individual's input.
  • No
    user-478855
    If used for correcting the style or grammar, I see no difference. Of course, I am an experimental scientist, and AI can not make up my experiments for me (although, it may help with ideas for experiments, but that is a different story)
  • Yes
    user-885754
    AI is prone to overestimating responses in its outputs.
  • Yes
    user-65167
    It is particularly useful for non native english speakers
  • Yes
    user-544906
    Faster and 24x7 available
  • Yes
    user-583633
    The use of AI in writing an article makes the process faster, more efficient and possibly more accurate.
  • Not sure
    user-764272
    Yes and no ... It depends how it is used. If the author gives ai the content to rewrite in a format that works for publications then it's exactly like other assistive tools. If it's asked to create something from anything it has access to then it could create something questionable.
  • Yes
    user-887652
    AI (typically in the form of LLMs) is a heap of word association occurrences being sorted very quickly. It is imitation, not writing. 
  • Not sure
    user-284769
    AI may be helpful as a tool, although quite different to, say, Word. The latter is little different from a pen in its essence, whereas AI pretends to be 'creative'. Hired individuals is another story. If a 'scientist' is incapable to write his publication he must not be among the authors and better pursue artisan endeavours. 
  • Yes
    user-765807
    Utilizing AI to assist in writing science publications does have some differences compared to using other tools or hiring individuals. Let’s explore these distinctions:
    1. Automation and Scale: AI can automate tasks at a larger scale than individual assistants. It can process vast amounts of data, generate summaries, and even draft sections of manuscripts. This efficiency is challenging to achieve manually.
    2. Objective vs. Subjective: AI tools operate objectively based on algorithms and patterns. Human assistants may introduce subjectivity, personal biases, or interpretations. Researchers need to be aware of this distinction.
    3. Learning and Adaptation: AI systems can learn from existing literature and adapt over time. They improve with more data and feedback. In contrast, human assistants have fixed knowledge and skills.
    4. Ethical Considerations: AI raises ethical questions related to transparency, accountability, and bias. Researchers must ensure that AI-generated content aligns with ethical standards.
    5. Collaboration: AI can complement human efforts. Researchers can collaborate with AI tools to enhance productivity and quality.
  • Yes
    user-35749
    It provide ideas 
  • No
    user-232098
    Although, its similar with other assistive tools but it differs significantly in ethical considerations, the potential for bias, and the necessity for transparency and validation
  • No
    user-634057
    Language models can't really create content without substance, so assuming authors have real research substance to publish, the language model in this case would only facilitate writing or telling the story of the results using a specific style as prompted by the authors.
  • Yes
    user-952116
    Used without critical thinking, AI can create a whole publication from scratch - no other tool existing today can achieve that. In this sense it is more than a tool and becomes more like a virtual collaborator.
0
user-911600
06/05/2024 13:26
AI will be useful for pathologists publishing results of scientific studies on respiratory tract toxicity.
0
user-911600
06/05/2024 13:27
AI will supplement but not replace human knowledge and experience. 
0
user-646369
06/06/2024 00:00
AI represents a tool, and not much more than that
0
user-809947
06/07/2024 09:09
AI is an engine that uses experiance and data gained by humans. It can be categorized as more advanced statistic tool
Please log in to comment.