Welcome to our new post! In this article, we present some curiosities about artificial intelligence and Chat GPT-4. We will address the following aspects: What is artificial intelligence?, How and when did it emerge?, advantages and disadvantages, the new functions of the Chat GPT-4, examples of AIs in our day and do AIs represent a threat to workers? Get ready to learn amazing things about this tool and surprise your friends with these interesting facts.
What is artificial intelligence?
Artificial intelligence is a combination of algorithms that aims to provide machines with the same capabilities as a human being.
How and when did it come about?
Alan Turing is the father of artificial intelligence and the forerunner of modern computing. Already in the 50s he affirmed that machines were capable of thinking: "If a machine behaves intelligently in all aspects, then it is intelligent."
What are the types of artificial intelligences?
The first type is the one who thinks like humans. These systems are mostly used for task automation, learning, document summarization, face recognition, and troubleshooting. As an example of this type of artificial intelligence we can cite artificial neural networks.
These networks use machine learning to interconnect data structured in layers that resemble the human brain. This technology allows artificial intelligences to learn from their mistakes and gradually improve.The second type that we will see in this video are AIs that act like humans. An example is robots.
The third type are systems capable of thinking rationally. In this system, an attempt is made for the machine to acquire the abilities to perceive, reason and act. An example is expert systems that are used to solve specific problems. And the last type of AI that we'll cover in this video is systems that act rationally. These systems are dedicated to imitating human behavior rationally, as an example we can cite intelligent agents.
Classifications of AI
There are two classifications of artificial intelligences: narrow and general. Those classified as narrow are limited to a single context, that is, they are capable of executing a task perfectly. Those classified as general already have a much broader employability.
Examples of narrow are: search engines, virtual assistants and image recognition applications. There is no AI classified as general, since a universal algorithm that was capable of perceiving, thinking and acting like a human being in any environment could not yet be written.
Examples of artificial intelligence
Did you know that you have been in daily contact with artificial intelligence for years and you had not realized it? Many mobiles allow you to unlock the phone screen just by recognizing your face, this is an example of AI that is very present in our lives. Virtual voice assistants are also artificial intelligences (Siri, Cortana and Alexa).
There are even AI-equipped drones to identify deforested areas and plant trees. Unmanned submarines capable of detecting pipeline leaks. AIs can also optimize traffic once you can avoid traffic jams and collisions.
For the health area there are chat bots that, depending on the symptoms you have, will tell you the possible diagnoses. Online stores have AIs that help them suggest the customer the right product according to their interests.
Advantages and disadvantages of artificial intelligence
Some of the advantages of AIs are: automating repetitive and tedious tasks, reducing human error, allowing humans more time to dedicate to creative tasks, increasing precision, and streamlining processes.
The disadvantages of artificial intelligences are: the data has to be reliable and up-to-date (otherwise it may give false information), there is a lack of qualified professionals for this technology to advance and the development of an AI is expensive.
What is the process for a machine to learn something?
Two types of training are usually used: machine learning (automatic learning) and deep learning. Machine learning feeds the machine data that can be labeled (supervised learning) or not (unsupervised learning).
Deep learning is based on imitating the neurons of the human brain and structures them in layers. Thanks to Big data technology that allows the storage, analysis and management of large amounts of data, AIs are advancing more every day.
What is Chat GPT?
It is an artificial intelligence trained to hold conversations. In addition to understanding your message, this AI is capable of replying to you coherently with several paragraphs. Some wrong information was reported in responses from the AI and even on the home page it puts a notice that the information may not be true. At the moment, the Chat GPT-3 does not replace the research stage. Requests can be made in English or any other language. By having a wide range of applicabilities, users frequently discover new ways of using this AI.
It can clarify a concept for you, make a summary about a topic, write a script or a book and even generate programming code for you. Anything that can be expressed in the form of language can be asked of him. If you don't understand an exercise or need to learn something, you can instruct the machine to help you learn instead of giving you the answers.
What's new in Chat GPT-4?
The new version was published on March 14, 2023. Chat GPT-4 will now be able to interpret images that users send to it. In an example taught by OpenAI the tool was able to generate a website based on a drawing.
You can also interpret and describe the context portrayed in the image. Images can only be inserted by users, that is, artificial intelligence does not generate images. In this version the AI responses have been optimized, now they are longer and more reliable. Although some users continue to report failures in the AI responses, both in logic and in information.
One of the implementations is that now you can be instructed to answer from the point of view of a specific character, instead of the classic tone, the messages can have a more personalized tone.
Version 4 can only be accessed at the moment if you have a Chat GPT Plus subscription which now has limited access, because the company is scaling the capacity of the servers. The free version will use version 3.5 and will not have these implementations yet.
Do AIs represent a threat to workers?
AIs are very good especially for people with low budgets and startups that can't afford to commission a professional, but I don't think they can completely replace it. In addition, it allows optimizing the work of professionals, once the machine is in charge of repetitive tasks. This allows humans more time to devote to creative tasks.
At the moment I think that the AI will not replace the workers, because you have to know very well what to ask for to do it the way you want it. They don't always return reliable answers or good-looking images. They still need a human to oversee their work. This technology is still new and developing, you need implementations to improve your data sources and logical thinking.
However, jobs that can be automated with scripts and that contain many repetitive tasks may be affected in the future, but right now not all companies can afford to implement this technology.
📚 Data source:
No comments:
Post a Comment