Comment les chatbots peuvent-ils améliorer la rédaction technique ?

Chatbots should be considered only as references to be sure nothing is omitted, so that one does not have to conduct a conventional search, one reference at a time. This check is best done after the first draft by the person responsible. The creator is the human, not the robot. Trying to follow the robot's lead is a relinquishment of creativity, motivation and dignity.

I have experimented with using chatbots for technical subjects. I find that they can save time by combining a search function with a writing/organizing function. For a busy or over-committed technical writer, this can be an advantage

People often say that AI may replace people one day but, what if we didn't look at them that way? What if we say them as tools and not competitors especially for research. As a technical writer, I am intrigued by the concept of chat bots and I am excited to see what else technology will birth.

Los chatbots entregan herramientas de trabajo para mejorar la fluidez y el entendimiento de temas complejos y con alta carga técnica. En cualquier caso, es responsabilidad de quien utiliza esta herramienta el revisar minuciosamente su respuesta y si la información que entrega se ajusta a la necesidad. Más que un ahorro de tiempo, los chatbots pueden ser un aliado para ampliar la visión a otras necesidades.

Speed and clarity is of essense in today’s world. Chatbots bring to bear this and therefore make life easier especially where one has to manage work, family and study.

Chatbots can write only in a specific way. The writings lack practicality and the usage of certain words are repetitive. They also fail to create that emotional connection with their works, which is something that is extremely necessary for any writer.

When I wrote my first book, I relied on Word's grammar advice. It never took long to discover it was terrible to read. Great grammar, just devoid of life. I decided to [mostly] ignore it. Write as I spoke and sometimes used grammar we were taught not to use. I thought not using it made the book much better. I suspect AI will lead to the same issues - flat and predictable. I have read some that reads like management speak I thought said little to nothing. I suspect AI will eventually develop a style - like a pop group - derived from common programming and "writers" asking for the same type of writing. I found the hardest part of writing was getting sections started... With AI, I fear entire chapters will be used to "start". Steve

Apesar de os chatbots otimizarem o tempo e os recursos do time, sua utilização encontra uma barreira relativamente complicada: o uso de respostas padronizadas e menos pessoais. Além disso, sua utilização em larga escala e para o atendimento de casos complexos podem demandar a revisão de outros profissionais a fim de garantir que as respostas oferecidas estejam sendo eficazes. Ou seja, minimizando algumas das vantagens em se usar essa ferramenta.

Chatbots learn from data, and sometimes that data can be messy. Inaccurate information or misinterpreted language can lead to technical docs with embarrassing errors, like a robot mistaking "install" for "uninstall" and causing chaos!

Model collapse occurs when new AI models are trained on AI-generated content. These new models can give erroneous answers and distort the original information. It's like a game of telephone between AIs. This is a major future concern, especially as people start flooding the internet and their company documentation with AI-generated content. Imagine a scenario where errors in technical documentation are propagated into future versions..without anyone noticing these tiny errors (think safety manuals, medical documentation, code documentation, etc). Imagine an AI agent trained to give real-time answers..but it was trained on AI-generated content. Now realize that there are other unforeseen consequences that we are completely unaware of.

Chatbot content can be adapted to tree levels of language. Creative, normal or specific. As the result we can adapt the text easily for the users, the futur readers. It looks like an opportunity to develop our creativity about how to write the prompt. That would give us time saved. Or a source of inspiration, as a starting point of writing.

I was a technical writer in the Clinton Navy that failed my first technical writing class on a med cruise in 1994. I was devastated to the point that I gave up the dream of writing for a decade. However, I took a world history class afterwards, not giving up education in general but concentrated on writing for myself documenting my travels on the cruise. I lost those journals in a move back to my hometown. My journals were so detailed with feelings associated with landmarks from my travels they are now part of my working mind. These allegories are part of my makeup in a story I often use in a conversational tone that springboard my talks with doctors and CEOs in LinkedIn Audio Events. A chatbot in formation entering a new frontier.

Engaging, informative, and always ready to pour a delicious conversation. Here's the secret recipe for crafting content that keeps users coming back for more..

Sabe o lide que é ensinado no curso de jornalismo? Aquelas perguntas vitais que, respondidas com categoria, podem situar o leitor? Então, é isso. O que, quando, onde, porque e como. Sempre busco em chatbots trazer essas indagações; e para além disso, reforço questões como tom de texto, palavras-chaves, pesquisas necessárias e conclusões que possuem um raciocínio singular.

Chatbots have a problem with hallucinations. This means they sometimes make stuff up. In technical writing, this could lead to a disaster. If you write something that turns out to be false, you will lose credibility. In some fields, a wrong concept could lead to an accident. Even if a chatbot gives you a source, you can't trust it! Always double check to make sure your information is accurate. Always.

It's too easy for organizations to cut corners with technical writing. Only organizations with a commitment to documentation, experienced staff, and business-driven use cases to augment their technical writing with chatbots should even attempt to do a proof of concept of such a solution While AI won't replace a writer, a short-sighted manager with AI can certainly replace that writer. AI is still at a stage in too many enterprises that can invite hype and self-styled innovators, and those people will most certainly target the technical writers if they can.

Though Chat bots can be quite helpful, you should be careful about using AI generated links and sources. I figured this out when I was searching for a source for a project I was working on. I was actually elated I finally found the materials for my work. When I double-checked the source though, it returned an empty search. Which means the source does not exist. Some chatbot facts and responses may be reliable, but you can never be too sure if you don't do your due diligence and perform a more in-depth research.

Éramos papel, lápis, borracha e ideia. Nada podia salvar os nossos erros. A gramática era vossa responsabilidade, assim como a imaginação. Na escola, eu vendia o pensamento apalavrado. Com o esforço à punho, pagava o pastel do bar do colégio. Era uma inteligência real, talvez não ideal, mas de estímulo para jamais desistir de pensar. Hoje talvez meus colegas não precisassem mais me pagar um pastel em troca de uma redação. Quiçá eu mesmo não enxergasse razão em escrever se não fosse um desafio. Aliás, talvez a minha geração escrevesse bem menos e pior. Estou nostálgico e com temor pelos pastéis do futuro, assumo. Mas, principalmente, pensando o que formaremos e se formaremos amanhã. Escrever é existir e não escrever o contrário disso.

When it comes to your content, you are the authority, not the chatbot. I like using chatbots to criticize my work and they often suggest "improvements" by automatically rewriting sections. Typically, they shorten my explanations which increases the complexity of the sentences since they use high-level vocabulary and sentence patterns. While the rewrites are impressive, many times they completely miss the point: the target audience. I know which concepts are hard to understand, so in these cases, I try to reduce the cognitive load of my audience: I don't want them to deal with complex sentences while they are learning complex concepts. Remember, chatbots are just clever language tools, so take their suggestions with a grain of salt.