It’s been a year already since the world was blown away by OpenAI ChatGPT. Suddenly, everyone realized the potential of a large language model (LLM) for conversational chat, text generation, and even serving as an AI tutor. If that wasn’t enough to disrupt the tech industry, the release of OpenAI GPT-4, a multimodal large language model, really set the rest of Silicon Valley on notice. Not to be outdone, it’s been an AI arms race ever since with Meta, Google, Amazon, Microsoft, and countless others vying for consumers and enterprises alike to adopt their generative AI platforms.
With the end of 2023 quickly approaching, let’s look at the year’s top generative AI large language models.
What is a Large Language Model?
An LLM is a type of artificial intelligence (AI) that has been trained on a massive dataset of text and code. This training allows LLMs to perform a wide range of tasks, including text generation, translation, question answering, and code generation.
LLMs are typically trained using a type of machine learning called deep learning. Deep learning models are made up of many layers of artificial neurons, which are inspired by the neurons in the human brain. Each layer of neurons learns to extract different features from the data, and the features learned by each layer are combined to produce the final output of the model.
LLMs are still under development, but they have the potential to revolutionize the way we interact with computers. They can be used to create new applications in a variety of fields, including healthcare, education, and customer service.
The Top Large Language Models of 2023
Generative pre-trained transformers are foundational components of every large language model. While they have been in development for years, 2023 represented the wide adoption and embrace amongst the tech industry for generative AI. Here’s a look at some of the top large language models of 2023 for consumer and enterprise adoption:
- GPT-4 by OpenAI: This is the latest and most advanced multimodal large language model from OpenAI. Debuting in March 2023, it reportedly has 1.7 trillion parameters and can perform a wide range of tasks, including text generation, translation, and even tutoring. GPT-4 scored in the 90th percentile on the Uniform Bar Exam and the 99th percentile for the Biology Olympiad. That put the whole world on notice: Generative AI is for real.
- OPT-175B by Meta: Meta’s flagship LLM is similar in size and capabilities to GPT-4 with over 100 billion parameters. It is also open-source, which means that anyone can use it for free. For transparency, Meta is also providing the datasets it is trained on.
- Llama 2 by Meta: This release includes model weights and starting code for pre-trained and fine-tuned Llama language models (Llama Chat, Code Llama), ranging from 7B to 70B parameters. Llama 2 has double the context of Llama 1 and is open-source and available for download.
- LaMDA by Google: While LaMDA was originally released in 2021, it’s still widely utilized by consumers and enterprises alike. The first version of Google Bard was based on LaMDA, and subsequent updates to Bard in October 2023 changed the foundational model to PaLM2.
- PaLM2 by Google: This is one of the largest LLMs in the world, with 340 billion parameters trained on 3.6 trillion tokens. It can perform a wide range of tasks, including text generation, translation, and question-answering. Google has announced betas of specialized variants of PaLM2 for cybersecurity (Sec-PaLM2) and medical (Med-PaLM2) use.
- Claude by Anthropic: Anthropic is quickly rising, and Claude is a large reason why. Claude can help with use cases including summarization, search, creative and collaborative writing, Q&A, coding, and more.
While there are countless other LLMs in existence, the above are the most sought-after foundational models for research, enterprise, and consumer use alike. More consumer platforms and websites are quickly adopting generative AI, and there’s a good chance you’re actually interacting with a custom-tuned model from the above list. You just may be banned from using ChatGPT on your employer’s corporate network, at least.
