OpenAI has officially announced GPT-4 – the latest version of the incredibly popular large language model that powers artificial intelligence (AI) chatbots (among other exciting things).
If you’ve heard the hype about ChatGPT (perhaps at an incredibly hip party or work meeting), then you may have a passing familiarity with GPT-3 (and GPT-3.5, a more recent enhanced version). GPT stands for Generative Pre-trained Transformer, a machine learning technology that uses neural networks to bounce around raw input information like ping pong balls and transform it into something understandable and persuasive for humans. OpenAI claims that GPT-4 is its “most advanced artificial intelligence system” that has been “trained using human feedback to produce even safer, more useful results in natural language and code.”
GPT-3 and GPT-3.5 are large language models (LLM), a type of machine learning model, from AI research lab OpenAI and are the technology behind ChatGPT. If you’ve been following the recent developments in the AI chatbot arena, you’ve probably not missed the excitement about this technology and the explosive popularity of ChatGPT. Now, the successor to this technology, and possibly to ChatGPT itself, has been released.
Get to the point
- What is this? GPT-4 is the latest version of the large language model used in popular AI chatbots
- When does it come out; Officially announced on March 14, 2023
- How much? It’s free to try, and there are subscription levels
When will ChatGPT-4 be released?
GPT-4 was officially revealed on March 14th, although it wasn’t that much of a surprise as Microsoft Germany CTO Andreas Braun speaking at the AI in Focus – Digital Kickoff event let it slip that the release of GPT-4 was imminent.
It was previously speculated that GPT-4 would be multimodal, which Braun confirmed. GPT-3 is already one of the most impressive natural language processing (NLP) models, models built to produce human speech, in history.
GPT-4 will be the most ambitious NLP we’ve seen, as it will be the largest language model in existence.
What is the difference between GPT-3 and GPT-4?
The input type of the Chat GPT processes (iGPT-3 and GPT-3.5) is plain text, and the output it can produce is text and natural language code. The multimodality of GPT-4 means that you may be able to input different kinds of input – such as video, audio (eg speech), images and text. Like its capabilities at the input end, these multimodal capabilities will also enable the creation of output such as video, audio and other types of content. The input and output of text and visual content could give a huge boost to the power and capabilities of AI chatbots based on ChatGPT-4.
Additionally, similar to how GPT-3.5 was an improvement on GPT-3’s conversational capabilities, being more optimized for natural conversation, the ability to edit and extract code, and perform traditional integration tasks, GPT-4 will should be an improvement on Understanding GPT-3.5. One of the main advantages of GPT-3/GPT-3.5 is that they are trained on a huge amount of text data coming from the Internet.
What can GPT-4 do?
GPT-4 is trained on a diverse range of multimodal information. This means that, in theory, it will be able to understand and produce language that is more likely to be accurate and relevant to what is being asked of it. This will be another notable improvement in the GPT series for understanding and interpreting not only the input data, but also the context within which it is placed. Additionally, GPT-4 will have an increased ability to multitask.
OpenAI also claims that GPT-4 is 40% more likely to provide factual answers, which is encouraging to know as companies like Microsoft plan to use GPT-4 in search engines and other tools we rely on for factual information . OpenAI also said it is 82% less willing to respond to requests for “prohibited” content.
Security is a big feature of GPT-4, with OpenAI working for over six months to ensure it is secure. They did this through an improved monitoring framework and by working with experts in various sensitive fields, such as medicine and geopolitics, to ensure that the answers it provides are accurate and secure.
These new features promise greater ability and scope to perform a wider variety of tasks, greater efficiency of processing resources, the ability to complete multiple tasks simultaneously, and the ability to be more accurate, something that concerns current AI-bot and search engine engineers.
How GPT-4 will be presented is yet to be confirmed as there is still much to be revealed by OpenAI. We do know, however, that Microsoft has exclusive rights to OpenAI’s GPT-3 language model technology and has already begun full development of its ChatGPT integration into Bing. This leads many in the industry to predict that GPT-4 will also end up being integrated into Microsoft products (including Bing).
We have already seen the widespread and persistent ripples caused by GPT-3/GPT-3.5 and ChatGPT in many areas of our lives, including but not limited to technology such as content creation, education and commercial productivity and activity. When you add more dimensions to the type of input that can be submitted and generated, it’s hard to predict the scale of the next twist.
Ethical debates about AI-generated content have proliferated as quickly as the technology’s ability to create content, and this development is no exception.
GPT-4 is far from perfect, as OpenAI admits. It still has limitations around social biases – the company warns that it could reflect harmful stereotypes, and it still has what the company calls “illusions”, where the model creates fabricated information that is “wrong but sounds plausible”.
Even so, it’s an exciting milestone for GPT in particular and AI in general, and the pace at which GPT has evolved since its launch last year is incredibly impressive.