ChatGPT and artificial content creation

What is ChatGPT and what is artificial content creation?

Recently there’s been a lot of talk about the major revolution that artificial intelligence which can generate texts, transcriptions, or even write programming codes can be.

But do you know exactly what ChatGPT is and how it works?

ChatGPT is a free chat, albeit with a paid service also available, developed by the company OpenAI, but it isn’t the first free chat with artificial intelligence, nor will it be the last. Even just a quick Google, PlayStore or AppStore search will reveal an almost endless catalogue of software and apps before and after ChatGPT. YouChat, ChatSonic, Perplexity AI and Jasper Chat are some of the names you’ll find in this list.

Why has ChatGPT managed to stand out from the crowd?

On 30th November 2022, OpenAi launched the new version of ChatGPT and in just five days it reached one million users. Shortly afterwards, at the start of 2023, Microsoft confirmed its investment of ten billion dollars in the company that created ChatGPT (Open AI), extending the collaboration between the two companies which began in 2019 for the development of artificial intelligence.

Given this situation, Microsoft is positioned as a potential candidate to implement ChatGPT or its future versions in some of the products it already offers to its users.

While the creation of content through the use of artificial intelligence is a developing field of technology, it still has some limits that separate it from the human capacity to develop content of linguistic and cultural value.

The limits of Chat GPT:

1. Format limitations. The openAI ChatGPT format is, for the time being, text-only; it is not possible to add images, urls, videos or audio files. It can issue code which generates this type of file, but this makes its conversational capability somewhat limited. In addition to this, it is unable to read the intentionality of sarcastic, humorous or ironic text. It cannot take the cultural aspect of the language or the area where the sender of the message comes from into account.

2. Limitations on the length of responses. Chat GPT cannot write novels, essays or plays, just as the user is not allowed to enter excessively long texts in the chat for a query or request. The reason for this is, in terms of ChatGPT’s responses, the response time. From the user’s perspective, the number of variables and requirements which can be inserted in a long message or request.

3. Obsolete data. When querying ChatGPT, the most up-to-date data it can offer, for the time being, is from September 2021. Given the time since that last publication of almost more than a year, it can be said that the chat is operating from an obsolete dataset.

4. It is incapable of making value judgements. As a language model, ChatGPT cannot make value judgements. It aims to provide accurate and relevant information, based on data and facts about what is suggested and/or asked by users. From a neutral and impartial stance. This means that, faced with a situation with several options, the Chat will not choose any of them, but will instead respond by sidetracking around each option without coming to any definite conclusion about the question asked.

5. Memory problems. ChatGPT does not have a conversational memory which allows it to maintain a fluid interaction for a period of time, during which the user submits several edits or requests on the first response offered by the chat. This means that, if the user asks the chat for a text, to which they start requesting edits and changes, after a few interactions ChatGPT will lose the thread of the conversation and answer with something completely irrelevant.

6. Confidence in the data. ChatGPT is not connected to the internet. While a search engine is a system that searches for files and information stored on web servers, ChatGPT does not have this capability. The information it uses to respond is based on the data and conversations it has previously used as a training source. These include content up to the year 2021. This means that it does not have access to up-to-date information such as that you might find in a browser or consult with a voice assistant such as GoogleHome or Alexa. Likewise, as stated in the OpenAI FAQ, the conversations that any user has with the tool will be reviewed and stored in order to improve and expand ChatGPT’s linguistic resources and knowledge.

Concern about possible breach of European data protection rules (GDPR)

EU data protection rules represent the guidelines by which citizens can exercise greater control over the use of their personal data and businesses can enjoy a level playing field.

Information on this regulation, for both businesses and individuals, is available on the European Commission website.

Given the scope of knowledge that ChatGPT has access to, despite not being connected to the Internet, its use by businesses and individuals has resulted in concerns about compliance with these data protection rules.

In Italy, the Data Protection Authority (DPA) has already banned the use of this tool.  The regulator claims that OpenAI has no legal basis to support and justify the collection and storage of personal or business data for the development of ChatGPT’s algorithms.

But Italy, although the first country to take this measure, will not be the last.

Spain is another country which has already posited a ban on this tool for the same reasons as Italy. The Spanish Data Protection Agency (AEPD) has already submitted a request to the EU’s Data Protection Committee to assess the impact of the use of ChatGPT on data protection.

Other countries such as France, Germany and Ireland are also considering following Italy’s lead.

On the other hand, the US and Chinese governments have begun to evaluate possible regulations or barriers which could be applied to the use of ChatGPT. The aim of these would be to introduce a possible regulation of the tool’s use.

Technological evolution is important and helpful to society, but we don’t leave our work and its value in the hands of mere machines. The quality assurance seal of the human contribution is an unmatched and unique line which is instilled in each project.

At BigTranslation we defend the unique value of translation, localisation and content creation by professional linguists who are native speakers of their working language. With personalised attention to the needs of each of our clients.

Our translation and content creation providers are real people, industry professionals with expertise in linguistic specialities. Furthermore, we only work with native translators, professionals who know their language and culture.

👉      We’ll provide original, high-quality content with the assurance of a native speaker to your                              website. You can place your order or calculate your quote without any obligation.