Making artificial intelligence human-like is one of the most amazing scientific breakthroughs in science, but can it really replace human intelligence?
Artificial intelligence is an interdisciplinary technology that combines computer science, cognitive knowledge, linguistics, psychology, neuroscience and mathematics. With this commitment, scientific advances have sought to create machines that imitate human behavior and can carry out activities that make work more agile and efficient.
“This is just the beginning of the golden age and it comes with advances like GPT 3 , a deep learning tool that understands and processes human language without programming. This is a technology similar to IBM’s Deepblue , the supercomputer for playing chess, or AlphaGo , Google’s DeepMind artificial intelligence program,” according to Rita Estevez Luaña, CEO for Spain and Portugal at Experian, a leading multinational in data and analytics. advanced.
The current technological advances open up the possibility to continue exploring in the field of science and to have robots that are more and more similar to humans in the future. In the specific case of OpenAI’s Generative Pretrained Transfomer -GPT 3 , it was a digital tool programmed with public books available online, scientific papers and web research from all fields.
“Everything we value in the real world is based on our intelligence. And if we had access to much more, there would be no limits to what the human race can create”, Stuart Russell, researcher in artificial intelligence.
“Once the machine analyzed all the information, it created connections on a 700 GB model located on 48 16 GB GPUs. The previous version had 1,500 million parameters and the current one has 175,000 million ”, according to an article on the web portal Blog thinking big.
This is a technology that significantly revolutionized the industry and allows human intelligence to continue to understand cognitive and complete operations that continue to create evolutionary intelligence.
How will artificial intelligence advance in the future?
According to studies by the research center of the University of Oxford, Future of Humanity Institute , these are some of the main advances.
- By 2027, the machines will be able to drive trucks and this seeks to reduce the accident rate on roads.
- In 2049 AIs will be able to write Best Sellers and work in operating rooms to perform surgeries.
- It is expected that by 2057 they will be able to carry out scientific research.
- By 2024, artificial intelligence will overtake humans in language translation.
Human skills to work with artificial intelligence
- Develop emotional intelligence
The balance between logic and emotions is key to improving the development of digital tools and human work. In this way, communication and empathy are enhanced.
- Work with methodologies
Methodologies that place people at the center of production processes are key to creating synergies between digital tools and people.
- critical thinking
Although artificial intelligence is capable of making logical connections, it does not think for human beings, nor does it detect variables that go beyond the order of the programmed code.
“ Although artificial intelligence is capable of multitasking, solving complex problems in a wide range and using different technologies at the same time, it must always be controlled by a human who programs it and gives it orders”, says Ramón López de Mántaras, director from the Artificial Intelligence Research Institute of the CSIC in Barcelona.
- Multi-purpose chatbots : One of them will be the evolution of the GPT chat that will have images or videos with text.
- Automatic robots that calculate the carbon footprint : theartificial intelligence startup Hugging Face will advance its language model based on a technology that allows estimating the carbon footprint.
- Legal limits of AI: both the European Union and the United States Government are working on artificial intelligence regulations that limit some AI operations and in which companies are held responsible if they cause damage or privacy violations or unfair decisions made by algorithms.
Artificial intelligence is increasingly evolving to process information logically. However, the brain is evolutionary and each time it can improve its ability to analyze data and configure the machine to configure the information according to the orders it is given.
- Strong AI: this responds to the most futuristic technology where machines are capable of matching human intelligence.
Loose AI: It is a technology capable of solving defined and bounded reasoning problems.