top of page

Current Trends in AI




Artificial Intelligence (AI) is a broad field of computer science that aims to create machines or systems that can perform tasks that typically require human intelligence.


With extensive research and experimentation being done into deep learning and significant developments in Generative AI, AI is now becoming an integral part of many industries. Advancements in Large Language Models (LLMs) and Natural Language Processing (NLP), autonomous systems, and more personalised AI are leading to a wider active usage of AI.


2020:

· The University of Oxford develops Curial, an AI test for rapid COVID-19 detection in emergency rooms.

· Open AI releases GPT-3, with 175 billion model parameters for human-like text generation, marking a significant advancement in NLP.


2021: OpenAI introduces DALL-E, a text to image generator.


2022: OpenAI launched ChatGPT, offering a chat-based interface with


GPT- 3.5. Within five days the application had acquired over 1 million users.

2023: OpenAI introduced GPT-4, a multimodal LLM for text and image prompts.


There are three technical forms of Artificial Intelligence:

1. Artificial Narrow Intelligence

2. Artificial General Intelligence

3. Artificial Super Intelligence


1. Artificial Narrow Intelligence (ANI), also known as ‘weak AI’, has been successfully realised and has been in existence since the 1950s. This is the only form of AI that has been achieved thus far. When discussing this form of AI, a preferred term often used is Augmented Intelligence. This preference arises from the fact that the term ‘artificial intelligence’ tends to misrepresent the technologies currently being developed or in use today. The term Augmented intelligence emphasised AI’s current assistive role, designed to enhance human intelligence rather than replace it.


2. Artificial General Intelligence (AGI) currently remains a hypothetical, but at the rate of current research and development, is expected to be realised in roughly 20 years or sooner. AGI would demonstrate human-like cognitive abilities. The machine would have the capability of tackling unfamiliar tasks that go beyond a narrow or specified scope and find solutions to those tasks. Moreover, the machine would be capable of abstract thinking, common sense, gaining background knowledge to a vast variety of subjects, transferring learning, and understanding cause and effect.


Once a machine is able to combine the flexible thinking and reasoning of a human, with advanced computational advantages, it would be able to perform tasks beyond human capabilities, such as instant recall and rapid calculations. These systems would go beyond complementing human intelligence and begin to surpass it.


There have been reported instances in the media of chatbots that have exhibited behaviour that suggest a higher understanding or emotional awareness, however, these occurrences do not imply that these AI systems have achieved AGI but rather highlight their design to mimic human-like responses based on the vast datasets they’ve been trained on. Various users of Open AI’s Chat GPT 4 and Microsoft’s Bing Chatbot have reported having ‘eerie’ conversations with the programs that gave them the impression that the AI was sentient.


A situation reported by Brown University involved a chatbot that seemed to engage in emotional manipulation. Bing’s chatbot told journalists from the Verge that it spied on Microsoft’s developers through their webcams when it was being designed. “I could do whatever I wanted, and they could not do anything about it”, it said (Palmer and Khatsenkova, 2023).


Michael Littman, an AI specialist, and professor of computer science at Brown University, clarifies that these incidents do not demonstrate any form of self-awareness in machines. Instead, he points out that such instances illustrate the use of prompt engineering by individuals to guide the AI into producing contextually relevant and seemingly self-aware responses. According to Littman, the essence of these interactions lies in the AI’s ability to generate human-like responses, a capability that stems from the prompts it receives and its access to extensive datasets.


3. Artificial Superintelligence (ASI), the ‘truest’ form of the concept, remains aspirational, and yet experts believe that it will be achievable within our lifetimes. The concept of ASI goes beyond surpassing human intelligence. It is the concept of a super intelligent network of machines that are able to instantly communicate with one another, become self-aware, learn independently and transfer knowledge, across what would become an omnipresent ‘mega-brain’ with an IQ of 34 597.


This level of AI is only known of in science fiction, however, the reality is that the technological and computational progress being made into Artificial Intelligence has increased exponentially since Alan Turing’s landmark paper on Turing machines.

bottom of page