(ORDO NEWS) — We used to call artificial intelligence a lot of programs in our devices. But in fact, real AI is much less common than you might think.
At the end of last month, artificial intelligence in the form of the ChatGPT neural network from science fiction and research labs hit our computers and phones.
This so-called “generative AI” can create an essay, or create a recipe and shopping list, and write, for example, a poem in the style of Elvis Presley.
Similar systems have already demonstrated even greater potential for creating new content, and the generation of images in a neural network based on a textual description simply blew up the Internet.
Real AI
AI may not yet have the living consciousness or theory of mind popular in sci-fi movies and novels, but it’s getting close to at least disrupting what we think AI systems can do.
Researchers working closely with these systems have swooned from the perspective of reasonableness, as in the case of Google’s LaMDA Large Language Model (LLM).
LLM is a model that has been trained to process and generate speech in the form of natural language. All this raises the question of how AI is still different from our intelligence?
To qualify as AI, a system must exhibit a certain level of learning and adaptation. For this reason, decision-making systems, automation and statistics are not artificial intelligence.
AI is broadly categorized into two categories: AI weak/narrow intelligence (ANI) and AI strong/general intelligence (AGI). To date, AGI does not exist.
A key challenge for creating strong artificial intelligence is to adequately model the world with full knowledge in a consistent manner. This is, to put it mildly, a massive undertaking.
Most of what we know today as artificial intelligence is actually weak AI – in this case, a particular system solves a particular problem.
Unlike human intelligence, such a narrow AI is only effective in the area in which it has been trained: for example, fraud detection, face recognition, or content recommendation.
AGI, however, will function just like humans. At the moment, the most notable example of attempts to achieve this is the use of neural networks and “deep learning” trained on huge amounts of data.
Neural networks are based on how the human brain works. Unlike most machine learning models that perform calculations on training data, neural networks work by passing each data point one by one through an interconnected network, adjusting the parameters each time.
As more and more data flows through the network, the parameters stabilize; the end result is a “trained” neural network, which can then produce the desired result based on new data – for example, recognize whether an image contains a cat or a dog.
A significant leap forward in the field of artificial intelligence today is driven by technological improvements in how we can train large neural networks, retuning a huge number of parameters in each run, thanks to the power of large cloud computing infrastructures.
For example, GPT-3 (the artificial intelligence system that runs ChatGPT) is a large neural network with 175 billion parameters.
—
Online:
Contact us: [email protected]
Our Standards, Terms of Use: Standard Terms And Conditions.