OpenAI has officially announced GPT-4 – the latest version of its incredibly popular large language model that powers artificial intelligence (AI) chatbots (among other cool stuff).
If you’ve heard the hype about ChatGPT (perhaps at an incredibly trendy party or work meeting), you may be a little familiar with GPT-3 (and GPT-3.5, a more recent enhanced version). GPT is the acronym for Generative Pre-trained Transformer, a machine learning technology that uses neural networks to bounce raw input information like ping pong balls and convert them into something understandable and convincing to humans. OpenAI claims that GPT-4 is the “most advanced AI system” that has been “trained using human feedback, to produce even more secure, useful output in natural language and code.”
GPT-3 and GPT-3.5 are large language models (LLM), a type of machine learning model, from the AI research lab OpenAI and they are the technology on which ChatGPT is built. If you’ve been following the recent developments in AI chatbots, you probably haven’t missed the excitement about this technology and ChatGPT’s exploding popularity. Now the successor to this technology, and possibly ChatGPT itself, has been released.
Cut to the chase
- What is it? GPT-4 is the latest version of the grand language model used in popular AI chatbots
- When is it out? It was officially announced on March 14, 2023
- How much is it? It’s free to try and there are subscription levels as well
When will ChatGPT-4 be released?
GPT-4 was officially unveiled on March 14, though it wasn’t much of a surprise as Microsoft Germany CTO Andreas Braun spoke at the AI in Focus – Digital Kickoff event, letting it know that the release of GPT-4 4 was at the door.
It was previously speculated that GPT-4 would be multimodal, which Braun also confirmed. GPT-3 is already one of the most impressive natural language processing (NLP) models, models built for the purpose of producing human speech, in history.
GPT-4 will be the most ambitious NLP we’ve seen to date, as it will be the largest language model in existence.
What is the difference between GPT-3 and GPT-4?
The type of input that Chat GPT (iGPT-3 and GPT-3.5) processes is plain text, and the output it can produce is natural language text and code. The multi-modality of GPT-4 means that you can potentially enter different types of input, such as video, sound (for example, speech), images, and text. Like the capabilities on the input side, these multimodal capabilities will also potentially enable the generation of outputs such as video, audio, and other types of content. Entering and executing both text and visual content can greatly boost the power and capabilities of AI chatbots that rely on ChatGPT-4.
Additionally, similar to how GPT-3.5 improved on GPT-3’s chat capabilities by being more refined for natural chat, the ability to process and execute code, and to perform traditional completion tasks, GPT would -4 should be an improvement over GPT-3.5 understanding. One of the main strengths of GPT-3/GPT-3.5 is that they are trained on a huge amount of text data from the Internet.
What can GPT-4 do?
GPT-4 is trained on a diverse spectrum of multimodal information. This means that, in theory, it will be able to understand and produce language that is likely to be more accurate and relevant to what is being asked of it. This will be another marked improvement in the GPT series to understand and interpret not only input data but also the context in which it is placed. In addition, GPT-4 has a greater capacity to perform multiple tasks simultaneously.
OpenAI also claims that GPT-4 is 40% more likely to provide factual answers, which is encouraging to learn as companies like Microsoft plan to use GPT-4 in search engines and other tools we rely on for factual information. OpenAI has also said it is 82% less likely to respond to requests for “disallowed” content.
Security is a big feature of GPT-4, with OpenAI working for over six months to ensure it’s 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 safe.
These new features promise greater power and range to perform a wider variety of tasks, greater efficiency of processing resources, the ability to run multiple tasks simultaneously, and the potential for greater accuracy, which is a concern among today’s AI bot and search engine engineers.
How GPT-4 will be presented is yet to be confirmed, as much remains to be revealed by OpenAI. However, we know that Microsoft has exclusive rights to OpenAI’s GPT-3 language model technology and has already started the full rollout of ChatGPT integration in Bing. This leads many in the industry to predict that GPT-4 will eventually also be embedded in Microsoft products (including Bing).
We have already seen the long lasting and sustained waves 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 both submitted and generated, it is difficult to predict the size of the next turnaround.
The ethical debates surrounding AI-generated content have grown as fast as the technology’s ability to generate content, and this development is no exception.
GPT-4 is far from perfect, as OpenAI admits. It still has limitations regarding social bias — the company warns it could reflect harmful stereotypes, and it still has what the company calls “hallucinations,” where the model creates fabricated information that “sounds false but plausible.”
Still, it’s an exciting milestone for GPT in particular and AI in general, and the pace at which GPT has progressed since its launch last year is incredibly impressive.