“Can machines think?” This question, asked by mathematician Alan Turing in the 1950s, dynamically changed the way we look at machines. In 1956 John McCarthy famously coined the term ‘Artificial Intelligence’ (AI) which described machines capable of performing tasks that typically require human intelligence. Over the years AI has been gaining popularity and often vendors tend to promote how their products and services utilize AI. In many cases, they are only referring to one aspect of AI such as machine learning. When in fact, AI is a combination of hardware and software which helps it write and train machine learning algorithms.
Generally, AI learns by devouring enormous amounts of data and deciphers the data for correlations and patterns. It uses these patterns to make predictions. For instance, chatbots are fed a blueprint of text chats that help them determine how humans interact with each other. Similarly, image recognition tools can learn to pinpoint and label objects in images by reviewing a surplus of data. AI focuses on 3 cognitive skills: learning, reasoning, and self-correction. AI can be broadly divided into 2 types: Narrow AI and General AI. (Also, find our discussion of an exciting development in AI: GPT-3)
Artificial Narrow Intelligence (ANI) also known as ‘Weak’ AI is the pre-existing AI in this world. Narrow AI is a 'soldier' that can only concentrate on one particular task like YouTube recommendations, the Spotify playlist, or driving a car. ANI can perform tasks in real-time but they draw their experience from a specific set of data. The outcome is that ANI cannot fathom performing a single task outside its comfort zone. Every machine that encircles us is Narrow AI. Google Assistant, Google Translate, Siri, and other natural language processing (NLP) tools are specimens of Narrow AI. The assumption why ANI is called ‘weak’ is that interaction with humans is unfulfilling. But the actual reason why ANI is considered weak is that it is machine lightyears away from human intelligence. It lacks the self-awareness, consciousness, and genuine intelligence to match us. ANI cannot think for itself. Human beings have the affinity to gauge their surroundings, to be sentient creatures, and to have emotionally driven responses to circumstances. The AI that exists surrounds us doesn’t have the variability or flexibility to think as we do.
But ANI should not be taken lightly. It is a great accomplishment of human innovation and intelligence. ANI systems can process information and complete tasks significantly faster than any human being: It takes care of the mundane that tire us. ANI has increased the quality of life by taking care of trivial tasks like asking Siri to take care of the shopping or guzzling through pyramids of data and analyzing it to make sense.
General Artificial Intelligence (AGI) is a machine that can comprehend data like a human but with an affinity to process that information at an unfathomable speed. If ANI is a soldier then AGI is more like a commander. AGI describes the concept of a sophisticated machine, not unlike the sci-fi movies and games in existence such as: Detroit-Become Human, Horizon: Zero Dawn or the movie Her. These examples showcase a futuristic world in which humans interact with machines that are conscious, sentient, and driven by emotions and are self-aware.
As much as one would like to expect humanoid robots walking among us in 2021, scientists and researchers have only scratched the surface of AGI. In order to achieve human-like consciousness, emotions, and flexibility—scientists would need to cultivate a machine similar to the human brain. The human mind is unfortunately still much of a mystery to this date. Theoretically, one can create algorithms that can replicate the complex computational abilities of the human brain. In simple terms, given infinite time and memory, any issue can be solved algorithmically.
So will we be seeing AGI walking around in the near future? It is not surprising to see what an enormously challenging task to achieve AGI is. While theoretically, one can possibly replicate the functioning of a human brain, it is far from being achieved practically. However, all is not lost. The rapid growth and development of AI are clearly visible. With AI developing new capabilities, scientists might surprise us with the development of artificial general intelligence. In conclusion, though we are far from achieving artificial general intelligence, the exponential advancement of AI research may culminate in the invention of artificial general intelligence within our lifetime or by the end of this century.
Want to learn more? We also have a discussion on the difference between AI and NLP.
We [Consensus] had a very complex and specific set of annotation needs. Datasaur was able to address those needs efficiently and effectively.
Information labeling tasks has been reduced by 80% which has allowed us to optimize our workflow much more, allowing us to focus on other areas that are also priorities for us.
"We looked at Prodigy, LightTag, LabelBox, Scale and more. You really can't beat Datasaur for their suite of features and price point."