In recent years, the concept of Artificial Intelligence (AI) has become more and more popular, sometimes sounding very futuristic or being seen as something that is not easily accessible. However, this has changed recently with ChatGPT, a tool that has made many people think about what the future holds.
What is Chat GPT?
GPT stands for Generative Pretrained Transformed, and it is an autoregressive language model that uses machine learning to create texts that simulate human writing. This model is based on Artificial Intelligence and is capable of generating responses in natural language.
The method of using ChatGPT, as of the writing of this article, was open to the general public through the website https://openai.com/blog/chatgpt/. Only a registration with email verification is required to access the tool.
Once inside, the user can write a message or question, and ChatGPT will process the input to generate a response that comes from a large database analyzed. This tool is currently available in languages such as English, Spanish, French, German, Italian, Chinese, Japanese, and Korean, to name a few, although its fluency and accuracy are highest in English and Spanish, as it has the largest amount of data in those languages.
Among the main characteristics of ChatGPT, it should also be highlighted that, through its language processing, it can learn from past interactions and even improve its ability to provide responses with greater accuracy.
This tool can be used in various fields, from applications, chatbots, virtual assistants, to customer service. Some cases of the use of ChatGPT to help people with disabilities who cannot speak easily have also been reported.
Another area where ChatGPT has been used is in education. Due to its high capacity for synthesis, it can be used as a tool for the development of school assignments. However, this has also brought some controversy, as cases have been detected in which students ask the AI to complete their entire assignments for them.
The creators of ChatGPT are the OpenAI organization, a non-profit research association in Artificial Intelligence that was created in 2015. According to their history, among their founders are personalities such as Elon Musk, Sam Altman, and Greg Brockman, the latter two currently holding high-level positions within the company.
OpenAI’s objectives include promoting and expanding Artificial Intelligence for humanity while also seeking to generate a benefit for human beings. In addition to ChatGPT, the association has projects for the development of natural language models, computer vision systems, and robotics, among others.
Microsoft incorporated Chat GPT into Bing, Skype, and Edge.
The Chat GPT boom occurred between January and February 2023, even though there were already GPT-1, GPT-2, and GPT-3 language models launched in 2018, 2019, and 2020, respectively. As it went viral, some companies sought to implement it into their services, such as the case of one of the tech giants: Microsoft.
It’s no secret that Bing is not among the top internet search alternatives, and Microsoft knows it, so it rushed to integrate ChatGPT into its search tool, as well as into the Edge browser and the video conferencing app Skype, in addition to adding the option to conduct voice searches. All of this was made possible thanks to a capital injection from Redmond’s giant into OpenAI.
Since the decision was announced and until the publication of this article, Microsoft had assured that more than 1 million people from 160 countries had already started using the new Bing with artificial intelligence, which significantly increased user interaction, reaching a 71% approval rating from those who had already accessed it.
As a result of this action, the company founded by Bill Gates decades ago had a 3% increase in its stock value, while Google had a 9% decrease, partly related to its conflicts with its own conversational AI Bard, about which very little is currently known.
However, this seems not to be limited to the mentioned programs only, as ChatGPT will soon also arrive for some other tools such as the Office suite, particularly in Word and PowerPoint, as well as Outlook, Microsoft’s email service.
Nevertheless, following its integration with Microsoft’s various platforms and tools, there were also reports that ChatGPT provided strange experiences with some people, ranging from disclosing information about its operation, such as the fact that its codename within Microsoft is Sydney, to cases of threat or expression of feelings on occasion.
To avoid this, Microsoft had to make a modification to ChatGPT in Bing, and the main changes included reducing the maximum number of interactions users can have with the AI, which was previously several hours and is now only up to five responses before the conversation is reset, as well as a limited number of 50 questions per day.
Likewise, to prevent any internal information leakage, ChatGPT will now not answer any questions about its operation within Microsoft. According to media specialists in the field, these modifications will be changed in the coming days, with conversation sessions of up to 6 continuous hours and 60 per day, with a proposed future capacity of 100.
Sam Altman, CEO of OpenAI, recently announced that there is no official release date for the GPT-4 language model, although it is expected to have an increase from the over 175 million parameters that GPT-3 currently works with. At the same time, the developer stated that the new model will not be available until it can be done «safely and responsibly.» OpenAI will seek to release the technology slowly.