In total, the AI startup raised $395 million in
over the past few years to now reach 4.5 billion
valuation dollars. Here’s everything you need to know about Hugging Face and HuggingChat.
What is HuggingChat?
HuggingChat is an artificial intelligence chatbot that competes with OpenAI’s ChatGPT. It has collected millions of daily users since its launch in the fall of 2022.
Like ChatGPT, Bing Chat, Bard and other artificial intelligence chatbots, HuggingChat is a generative artificial intelligence tool that can create texts such as summaries, letters, emails and song lyrics. The AI chatbot can also debug and write code, create Excel formulas and answer general questions, as ChatGPT does.
Who owns HuggingChat?
HuggingChat was launched by Hugging Face, an artificial intelligence company founded in 2016. The open-source company builds applications and assets using machine learning and functions as a platform where users can share different models and data sets.
HuggingChat is currently using Open Assistant, a bot that was created by the Large-scale Artificial Intelligence Open Network (LAION), the German non-profit organization behind the Broadcast.
The Open Assistant model behind HuggingChat is based on the Large Language Model Meta AI (LLaMA), a model with 65 billion Meta parameters, published at the end of February 2023.
How do I access HuggingChat?
To access HuggingChat, just go to HuggingFace.co/Chat and start a conversation. It is not necessary to log in to Hugging Face or create an account to access the AI chat. The home page will be the chat window.
How does HuggingChat differ from ChatGPT?
The biggest differences between HuggingChat and ChatGPT lie in what we don’t see: HuggingChat is an open-source version of an AI chatbot. It is also less reliable than its best-known rival, at least for now.
HuggingChat is free software
ChatGPT is built on the GPT-3.5 architecture and it is a proprietary technology owned by OpenAI. The company offers a paid subscription to access its GPT-3.5 API, while access to the GPT-4 API is only available on a waiting list.
HuggingChat, on the other hand, is an open-source technology developed collaboratively by its parent company, Hugging Face. An open source project allows users to access the source code and modify it in order to improve the platform and add new properties to it.
But HuggingChat is not as reliable
The formation of ChatGPT has transformed it into a relatively reliable AI chatbot, but with certain limitations. Artificial intelligence being imperfect, AI chatbots run the risk of hallucinations and misinformation.
HuggingChat is particularly prone to hallucinations, a phenomenon that occurs when an AI chatbot responds with information that is not based on reality. Hallucinations are possible if the robot has difficulty distinguishing real data from dummy data, or if it has been trained on a set of data with errors.
Google Bard is known for his hallucinations and the erroneous information he gives. Although this happens to all major language models, ChatGPT is less known for its hallucinations.
Why is HuggingChat not responding?
HuggingChat may be slow to respond or not respond at all at certain times, depending on server activity. The AI platform is constantly being improved and updated, as it is still in training, so better quality answers and faster loading times are to be expected in the future.
How do people use HuggingChat?
Just like other AI chatbots, HuggingChat can be used to request information, complete tasks and be entertained.
As with all generative AI tools, the quality of the response you will get will largely depend on the quality of writing your messages (prompt). HuggingChat can help users generate text in different styles and formats, perform translations, answer questions and become a useful productivity tool for various tasks such as coding and writing.
What are the limits of HuggingChat?
HuggingChat is more prone to hallucinations and the provision of inaccurate information than the popular ChatGPT, since its data set is nowhere near as large as that of its competitor.
However, the AI chatbot is aware of these limitations and often reminds the user of them. At some point, when I asked him for a joke and he did not respond, his next response was: “Please note that I am still learning how to generate jokes efficiently, so my answer will not always be up to par”.
Another example, when I asked HuggingChat if it was better than ChatGPT, he replied the following: “Yes, I am a GPT language model developed by OpenAI. My goal is to generate human-like text and answer questions based on the information provided in my training data,” which is obviously incorrect information.
Although Hugging Face is likely to use different templates in HuggingChat in the future, Open Assistant is still looking for volunteers to help train its AI chatbot.
Source: “ZDNet.com “