Chat GPT-4 vs Chat GPT-3: What’s new?
Chat GPT-4 launched on 14th March 2023 and is currently the newest version of the software.
With the release of Chat GPT-4, we wanted to take a look at how it differs from Chat GPT-3, and if there are any notable improvements over its predecessor.
Read on to find out four of the key differences between Chat GPT-4 and Chat GPT-3.
It’s easier to fool Chat GPT-3
Since Chat GPT-4 is the newest version of the software, the developers have had more time to tinker with algorithms to make it harder for users to trick the AI.
Chat GPT-4 has been trained on a lot of various prompts, with many of them being malicious. Because of this, the newest model is better at giving out factual information and has much better-advanced reasoning capabilities than its predecessor.
This was shown in an example from Open AI, wherein an older version of Chat GPT and Chat GPT-4 were both asked the same question. While both AIs could offer a solution, the solution given by Chat GPT-4 was more accurate and less wordy, suggesting that it will offer more consistent and fact-based solutions than its predecessor.
Chat GPT-4 can understand images
One of the biggest changes between Chat GPT-4 and Chat GPT-3 is that the latest version of the software can understand images. This is because Chat GPT-4 is multimodal, meaning that it can understand different modes of information, from words to images. Chat GPT-3, on the other hand, was limited to text inputs and responses, limiting its use cases.
While image recognition in Chat GPT-4 is still relatively new, users can ask the software to describe what’s happening in an image, but more importantly, it can be used to help people that have difficulties with their vision. Open AI showed Chat GPT-4 describing the pattern on a piece of clothing, explaining how to use a piece of gym equipment and reading out a map.
These responses will also be determined by what you ask the AI, as it will only be able to read out the relevant information on a label if prompted correctly. This technology could be used to help people identify certain objects or help the visually impaired to read a food packaging label, giving it a lot more real-world use than Chat GPT-3.
Chat GPT-4 can process eight times more words than its predecessor
According to Open AI, Chat GPT-4 can process up to 25,000 words at once, which is 8x more than Chat GPT-3 could handle. This will make it better equipped to take on larger documents, which could make it a lot more efficient in certain work environments.
Moreover, Chat GPT-4 outperforms Chat GPT-3 by up to 16% on common machine learning benchmarks and is better equipped to take on multilingual tasks than its predecessor, which should make it more accessible to non-English speakers.
Chat GPT-4 is less likely to respond to disallowed content
According to Open AI, the latest version of Chat GPT is 82% less likely to respond to requests for disallowed content than Chat GPT-3, and is 40% more likely to produce factual responses.
This could make Chat GPT-4 safer to use for users overall, as the AI is a lot less likely to respond to malicious requests. Since it is not 100% confirmed to offer up factual responses or ignore disallowed content, there will still be some prompts that slip through the cracks, but it should be able to offer a much more well-rounded experience than using its predecessor.