Recently, there has been a lot of publicity about the planned upgrades to OpenAI’s ChatGPT AI-powered chatbot and Meta’s Llama system, which powers the company’s chatbots on Facebook and Instagram.
The technology behind these systems is known as a large language model (LLM). These are artificial neural networks, a type of AI designed to mimic the human brain. They can generate general-purpose text for chatbots and perform language processing tasks such as classifying concepts, analyzing data, and translating text.
They acquire these skills through an intensive process known as training, which exposes the AI system to vast amounts of data in an effort to improve what it does. OpenAI and Meta are expected to release the newer versions of their chatbots – called GPT-5 and Llama 3 respectively – before the end of summer 2024. But how will they differ from their predecessors and what value will they add?
Like its predecessor GPT-4, GPT-5 will be able to understand images and text. For example, users can ask it to describe an image, making it even more accessible to people with visual impairments.
However, GPT-5 will have superior multi-language capabilities, making it possible for non-English speakers to interface and interact with the system. This includes greater command of language translation. The upgrade will also have an improved ability to interpret the context of dialogue and interpret the nuances of language.
Compared to its predecessor, GPT-5 will have more advanced reasoning capabilities, meaning it will be able to analyze more complex data sets and perform more advanced problem solving. This reasoning will enable the AI system to make informed decisions by learning from new experiences.
Therefore, it can provide users with more accurate information. For example, the system’s enhanced analytical capabilities allow it to suggest possible medical conditions based on symptoms described by the user. GPT-5 can process up to 50,000 words at a time, which is twice as many as GPT-4, making it even better equipped to handle large documents.
It will include a higher level of emotional intelligence, allowing for more empathetic interactions with users. This can be useful in a variety of environments, including customer service. GPT-5 will also show a significant improvement in the accuracy of the way information is searched and retrieved, making it a more reliable source for learning.
It is said to go far beyond the functions of a typical search engine that finds and extracts relevant information from existing information sources, and aims to generate new content.
GPT-5 is also expected to exhibit higher levels of fairness and inclusivity in the content it generates due to additional efforts by OpenAI to reduce bias in the language model.
It will be able to communicate more intelligently with other devices and machines, including smart home systems. The GPT-5 should be able to analyze and interpret the data generated by these other machines and incorporate it into user responses. She can also learn from this with the aim of providing more customized solutions.
This could enable smarter environments at home and in the workplace. GPT-5 will be more compatible with what is known as the Internet of Things, where devices at home and elsewhere connect and share information. It should also help support the concept known as industry 5.0, where people and machines operate interactively in the same workplace.
GPT-5 will include more robust security protocols that make this version more robust against malicious use and misuse. It can be used to improve email security by allowing users to recognize potential data security breaches or phishing attempts.
Overall, OpenAI’s upgrade should be more versatile, more energy efficient in its calculations, and provide a more customizable and personalized service.
Meta’s Llama Upgrade
Llama-3 is Meta’s competitor to GPT-5. It features several improvements over its predecessor, Llama-2. It is a more capable model that will eventually have 400 billion parameters, compared to a maximum of 70 billion for its predecessor Llama-2. In machine learning, a parameter is a term that represents a variable in the AI system that can be adjusted during the training process to improve its ability to make accurate predictions.
Llama-3 will also be multimodal, meaning it can process and generate text, images and video. Therefore, it can take an image as input to provide a detailed description of the image content. Likewise, it can automatically create a new image that matches the user’s prompt or the text description.
It will be able to perform tasks in languages other than English and will have a larger context window than Llama 2. A context window reflects the range of text that the LLM can process at the time the information is generated. This means that the model can process larger chunks of text or data in a shorter time when it is asked to make predictions and generate responses.
Meta plans to launch Llama-3 in several versions to work with a variety of other applications, including Google Cloud. Meta has announced that more basic versions of Llama-3 will be rolled out soon, ahead of the release of the most advanced version, expected next summer.
The transition to this new generation of chatbots could not only revolutionize generative AI, but also mark the beginning of a new era in human-machine interaction that could transform industries and societies on a global scale. It will impact the way people work, learn, receive healthcare, and communicate with the world and each other. It will make companies and organizations more efficient and effective, more flexible to change and therefore more profitable.
This article is republished from The Conversation under a Creative Commons license. Read the original article.
Professor Abdul Sadka is the founder and director of the Aston Digital Futures Institute, a nationally leading center for research and innovation expertise, developing breakthrough generative AI solutions to automate industrial workflows and improve the quality of experience for customers.