Accessibility / Report Error

THE LANGUAGE MODELS IN HEALTHCARE AND THE ROLE OF CHATGPT: COMMENTS

OS MODELOS DE LINGUAGEM EM SAÚDE E O PAPEL DO CHATGPT: COMENTÁRIOS

ABSTRACT

The introduction of chatbots has been one of the most intriguing advances in artificial intelligence. There are numerous potential uses for artificial intelligence in clinical research. However, there are also some issues that require attention. Everyone agrees that AI requires a more stable foundation and that a cutting-edge approach is necessary for AI to operate effectively.

HEADINGS:
Health; Health facilities; Comprehensive healthcare; Artificial intelligence

RESUMO

A introdução de chatbots foi um dos avanços mais intrigantes da inteligência artificial. Existem inúmeros usos potenciais para a inteligência artificial (IA) na pesquisa clínica. No entanto, há também outras questões que requerem atenção. Todos concordam que a IA precisa de uma base mais estável. Todos podemos concordar que uma abordagem de ponta é necessária para que a IA opere de forma eficaz.

DESCRITORES:
Saúde; Instalações de saúde; Planejamento em saúde; Inteligência artificial

INTRODUCTION

Dear Editor, we found that the article on “FUTURE OF THE LANGUAGE MODELS IN HEALTHCARE: THE ROLE OF CHATGPT”, by Tustumi et al., assembles the greatest medical knowledge to develop an orientation for practice22 Tustumi F, Andreollo NA, Aguilar-Nascimento JE. Future of the language models in healthcare: the role of CHATGPT. Arq Bras Cir Dig. 2023;36:e1727. https://doi.org/10.1590/0102-672020230002e1727.
https://doi.org/10.1590/0102-67202023000...
.

The introduction of chatbots has been one of the most intriguing advances in artificial intelligence (AI), according to Tustumi et al.22 Tustumi F, Andreollo NA, Aguilar-Nascimento JE. Future of the language models in healthcare: the role of CHATGPT. Arq Bras Cir Dig. 2023;36:e1727. https://doi.org/10.1590/0102-672020230002e1727.
https://doi.org/10.1590/0102-67202023000...
. While there is no doubt that ChatGPT has the potential to transform the way healthcare is delivered, the authors came to the following conclusion: it is important to emphazise that it should not be utilized as a replacement for real healthcare experts22 Tustumi F, Andreollo NA, Aguilar-Nascimento JE. Future of the language models in healthcare: the role of CHATGPT. Arq Bras Cir Dig. 2023;36:e1727. https://doi.org/10.1590/0102-672020230002e1727.
https://doi.org/10.1590/0102-67202023000...
. Instead, they suggested that ChatGPT be viewed as a tool that may be employed to supplement and complement the work of these professionals, enabling them to give their patients a better treatment22 Tustumi F, Andreollo NA, Aguilar-Nascimento JE. Future of the language models in healthcare: the role of CHATGPT. Arq Bras Cir Dig. 2023;36:e1727. https://doi.org/10.1590/0102-672020230002e1727.
https://doi.org/10.1590/0102-67202023000...
.

There are numerous potential uses for AI in clinical research. On the other hand, there are some issues that require attention. The importance of accuracy and dependability has already been addressed. AI's potential to forecast the outcomes of clinical issues is the subject of research. It ought to be clear that using AI has some drawbacks. It should not create, analyze, or approve vital information without human review11 Kleebayoon A, Wiwanitkit V. Artificial intelligence, chatbots, plagiarism and basic honesty: comment. Cell Mol Bioeng. 2023;16(2):173-4. https://doi.org/10.1007/s12195-023-00759-x
https://doi.org/10.1007/s12195-023-00759...
. Besides, abuse may result from poor usage management strategies; for example, it might be used to automatically detect plagiarism and ghostwriting.

Undoubtedly, we all agree that AI requires a more stable foundation and that a cutting-edge approach is necessary for AI to operate effectively.

  • Financial source: none
  • Editorial Support: National Council for Scientific and Technological Development (CNPq).

REFERENCES

  • 1
    Kleebayoon A, Wiwanitkit V. Artificial intelligence, chatbots, plagiarism and basic honesty: comment. Cell Mol Bioeng. 2023;16(2):173-4. https://doi.org/10.1007/s12195-023-00759-x
    » https://doi.org/10.1007/s12195-023-00759-x
  • 2
    Tustumi F, Andreollo NA, Aguilar-Nascimento JE. Future of the language models in healthcare: the role of CHATGPT. Arq Bras Cir Dig. 2023;36:e1727. https://doi.org/10.1590/0102-672020230002e1727
    » https://doi.org/10.1590/0102-672020230002e1727

THE INDISPENSABLE HUMAN PARTICIPATION IN ARTIFICIAL INTELLIGENCE GENERATED DATA

A INDISPENSÁVEL PARTICIPAÇÃO HUMANA NOS DADOS GERADOS POR INTELIGÊNCIA ARTIFICIAL. CARTA RESPOSTA

Authorship SCIMAGO INSTITUTIONS RANKINGS

Dear colleagues,

Thanks for thFe Letter to the Editor:

The recent announcements of artificial intelligence (AI)-based programs, including ChatGPT (OpenAi), Bard (Google), and Co-Pilot (Microsoft), raised interest in their potential impact on medical assistance and patient care. Indeed, deep learning and other AI modalities have been proposed in several fields of medicine.

Generally speaking, AI strategies provide physicians with feedback to help diagnose and treat patients. Kleebayoon and Wiwanitkit, in a letter to our previously published manuscript33 Tustumi F, Andreollo NA, Aguilar-Nascimento JE. Future of the language models in healthcare: the role of CHATGPT. Arq Bras Cir Dig. 2023;36:e1727. https://doi.org/10.1590/0102-672020230002e1727.
https://doi.org/10.1590/0102-67202023000...
, outlined that AI should not make vital decisions without human review. We are glad about their interest in our manuscript and completely agree with them. AI should be used as a tool for healthcare providers, not as a substitute for them.

In reality, none of the companies that announce AI-based programs claim them to be a human substitute. Human participation is indispensable in AI-generated data (Figure 1). As in any conversation, the natural (or human-like) conversation programs demand a dual interaction. In the case of chatbots, humans need to interact with the machine to create a conversation that might fulfill human interests.

Figure 1
The indispensable human participation in artificial intelligence generated data.

Specifically in the medical field, the healthcare provider needs to direct the machine-human chat for the benefit of the patient. Consequently, there is no doubt that the healthcare provider needs a great deal of knowledge to build a relevant clinical issue. Besides, healthcare providers must have a critical assessment capability to identify chatbot misinformation and judge if the information applies to their patients.

Numerous studies proposed using AI in computer vision to interpret images, including the use of automation in the assessment of surgical skill with machine learning algorithms11 Lavanchy JL, Zindel J, Kirtac K, Twick I, Hosgor E, Candinas D, et al. Automation of surgical skill assessment using a three-stage machine learning algorithm. Sci Rep. 2021;11(1):5197. https://doi.org/10.1038/s41598-021-84295-6.
https://doi.org/10.1038/s41598-021-84295...
,22 Mascagni P, Vardazaryan A, Alapatt D, Urade T, Emre T, Fiorillo C, et al. Artificial intelligence for surgical safety: automatic assessment of the critical view of safety in laparoscopic cholecystectomy using deep learning. Ann Surg. 2022;275(5):955-61. https://doi.org/10.1097/SLA.0000000000004351.
https://doi.org/10.1097/SLA.000000000000...
. Naturally, surgeons with no capability of interpreting AI feedback or with a deficit in surgical skills are null, even supported by these AI programs. The surgeons and all healthcare providers should recognize that they are ultimately responsible for clinical decisions and patient outcomes. The AI is only a tool.

Lastly, AI is created based on a solid amount of information. Humans are responsible for creating new data, new research manuscripts, new methods, and so on, which are the base for any AI functioning. Humans are responsible for the development and improvement of AI, making it an increasingly more valuable tool for clinical practice.

  • Financial source: none
  • Editorial Support: National Council for Scientific and Technological Development (CNPq).

REFERENCES

  • 1
    Lavanchy JL, Zindel J, Kirtac K, Twick I, Hosgor E, Candinas D, et al. Automation of surgical skill assessment using a three-stage machine learning algorithm. Sci Rep. 2021;11(1):5197. https://doi.org/10.1038/s41598-021-84295-6
    » https://doi.org/10.1038/s41598-021-84295-6
  • 2
    Mascagni P, Vardazaryan A, Alapatt D, Urade T, Emre T, Fiorillo C, et al. Artificial intelligence for surgical safety: automatic assessment of the critical view of safety in laparoscopic cholecystectomy using deep learning. Ann Surg. 2022;275(5):955-61. https://doi.org/10.1097/SLA.0000000000004351
    » https://doi.org/10.1097/SLA.0000000000004351
  • 3
    Tustumi F, Andreollo NA, Aguilar-Nascimento JE. Future of the language models in healthcare: the role of CHATGPT. Arq Bras Cir Dig. 2023;36:e1727. https://doi.org/10.1590/0102-672020230002e1727
    » https://doi.org/10.1590/0102-672020230002e1727

Publication Dates

  • Publication in this collection
    15 Sept 2023
  • Date of issue
    2023

History

  • Received
    03 July 2023
  • Accepted
    14 July 2023
Colégio Brasileiro de Cirurgia Digestiva Av. Brigadeiro Luiz Antonio, 278 - 6° - Salas 10 e 11, 01318-901 São Paulo/SP Brasil, Tel.: (11) 3288-8174/3289-0741 - São Paulo - SP - Brazil
E-mail: revistaabcd@gmail.com