Many new concepts are emerging in this new stage of technological revolution. For example, what is artificial intelligence In general, it refers to the imitation of machines of the processes of the human mind, such as learning or reasoning.
However, there is one more layer to this concept. whatWhat is generative artificial intelligence? According to one note of TechTarget could be defined as a type of AI that produces various contents including text, image, audio or data.
This notion began to gain momentum when new public access platforms began to challenge people. A great example is ChatGPT. Who has not heard of the famous “oracle” that answers all your questions? Who has not even tried to talk with the bot with some curiosity?
Like these, many more: Bard from Google which is another conversational chat. EITHER midjourney from Microsoft in which you can generate images from AI. Or ChatPDF that analyzes and summarizes your documents in a matter of moments. In short, this concept is immersing itself in many dimensions of technologies.
This technology is not new. In fact, according to Target Tech, was introduced in 1960 with chatbots. However, it was in 2014 when he began to create images, videos or audios of real people. This was due to the introduction of Generative Adversarial Neural Networks, also known as GANs, which are a new way of using deep learning to generate life-like images.
How Generative AI works
The dynamics of generative AI is basically the generation of a response from an order/request that includes a text, image, audio or video. That is, I can ask ChatGPT to give me a summary of a newspaper article or I can ask Midjourney to generate a 60-year-old version of a photo of a young girl.
The interesting thing is that from a first response given by the generative AI platform, you can go shaping and guiding what you really want. For example, if I asked you for the summary of the newspaper article and you gave me a long one, I can request that it be only ten lines long. Or if she gives me the photo of the young girl as an adult and she doesn’t have the same color eyes, I can ask her to change them. Thus, it can be molded until reaching the final result.
What are your main uses of Generative AI
According to one note of Multiple AI Generative artificial intelligence has more than 70 uses or applications that vary between the creation of audio, images, videos or texts. Some of them are:
1- Imaging: Users can transform text into images.
2- image conversion: consists of modifying the external elements of an image such as color and tones, while preserving other essential ones.
3- Generation of 3D shapes: Although computer scientists are still working in this area, today shapes can be generated in a 3D model.
4- music creation: You can create melodies for an advertising campaign, for example. However, you have to be aware of copyright and where you get the music from.
5- text generation: You can formulate from dialogues to advertisements. Generally, this tool is widely used in marketing or communication industries.
6- personalized content: generates personalized text in reference to the needs of the person.
7- Code Generation: allows both tech and non-tech people to generate codes.
Conversational AI: a question and answer model: the user asks and the bot answers based on all the information in the system
These applications are the ones that concern people in a more “universal” way. Then there are other uses of generative AI that concern, for example, medicine, education, fashion, finance. You can read these uses in depth in the note of Multiple AI.
What are the concerns of Generative AI
As Yuval Harari, philosopher and thinker, rightly stated in a discussion With computer scientist Yann LeCun, the question of AI is in whose hands it is. That, then, is the first and foremost concern of the evolution of technology, since it can be used with malicious intent or even abused. For this reason, regulation is being put on the table today, so as not to reach a point of no return.
According to TechTargetsome of the specific concerns are:
1- You can provide incomplete information or inaccurate information.
2- It can cause new forms of plagiarism.
3- It is easier to create fake news.
4- We will not be able to distinguish between a real photo or one created by the AI.
This content was originally published in RED/ACCIÓN and is republished as part of the ‘Human Journalism’ program, an alliance for quality journalism between RÍO NEGRO and DRAFTING.
Supporting quality journalism is essential to maintain an informed society and build a solid democracy.