Can A Bigger Version of GPT-3 or Deep Mind’s Gato Take Us to AGI?

How far are we at achieving AGO with GPT-3 and the new Gato?

Tech enthusiasts around the world are not satisfied with advanced robotics as they're running fast towards the discovery of machines that would practically be like humans. As if seven billion humans are not enough! Still the taste of creation addicts the researchers and radiologists to continue this near-impossible task of creating secondary humans. This is where the concept of AGI or artificial general intelligence comes to play. An Artificial General Intelligence (AGI) would be a machine capable of understanding the world as well as any human, and with the same capacity to learn how to carry out a huge range of tasks. Every year some hyped headlines in media channels keep screaming eureka and declare that AGI is here, but in truth, it always turns out to be around the corner and out of reach. The invention of the latest Gato by Deepmind has followed the same path.

What's Brewing over AGI?

DeepMind recently released a research paper and published a blog post on its new multi-modal AI system. Dubbed 'Gato,' the system is capable of performing hundreds of different tasks ranging from controlling a robot arm to writing poetry. The company's dubbed it a "generalist" system but hadn't gone so far as to say it was in any way capable of general intelligence, you can learn more about what that means here. It's easy to confuse something like Gato with AGI. The difference, however, is that a general intelligence could learn to do new things without prior training. What they're preaching is complex: if you scale a deep learning-based system large enough, feed it enough data, increase the number of parameters it operates with by factors, and create better algorithms, artificial general intelligence will emerge.

It Seems Like We are Running in Circle

The researchers keep hoping that a computer capable of human-level intelligence will explode into existence from the flames of AI as a natural byproduct of the clever application of more power. But as days pass by, this theory is circling over the basic concept of feeding the AI machine with enough data. However, the bigger the data storage capacity grows it would never go close to human-like intuition or general intelligence. Both OpenAI and DeepMind are working on the AGI through GPT-3 and Gato for a long period of time. But these companies are not capable of addressing the first problem in solving complex problems of AGI that includes artificial intelligence models learning new things without any training data.

Nope! AGI is Still not Here

Gato can consist of the potential to perform better in the consumer market but there are no viable entry points for AGI. AGI does not need the pre-trained data for learning new things. But one can reap the benefits of Gato — playing 600 different games on a video game console. GPT-3 and Gato both need hard filters to eradicate faults and flaws like bias, racism, and abusive language to put the outcome consistently. Meanwhile, AGI is known for powering intelligent machines to mimic human tasks— understand, learn, as well as perform intellectual tasks. It can study the human mind and solve any complex problem with cognitive computing functionalities.

More Trending Stories