Rumour has it that The large-language model could be unveiled at an event on March 16
There have been a few leaks and talks that perhaps Microsoft might be showing more information at their AI event. We may even see GPT-4 which might just included video features, it’s got a lot of people excited.
Let’s face it OpenAI’s ChatGPT (chatbot) has taken the world by storm. Chances are if you read the papers or online news you will have seen it mentioned at least once, weather that is negative or postitive. So just imagine a more powerful version of this? rumour has it they might also be including text to video too!
So Who Let the Cat out of the bag?
Well we have to point our finger at Microsoft Germany’s CTO Andreas Braun, who made the announcement at an event on Thursday (March 9th). We are unsure if this is 100% happening but we wouldn’t be surprised due to the speed that the AI space is going at. GPT-4 will be the next large-language modal that OpenAI’s has released in it’s series. The current version is GPT-3.5 which already is quite impressive.
Braun claims GPT-4 will boast multimodal capabilities, which would be quite impressive. Having the ability to turn text into video would also be a big plus, other tools do use this technology but it would be nice to see how GPT-4 does this. It makes sense as other big companies (Meta and Google) are currently working on perfecting this.
It seems that OpenAI’s has been keeping their cards close to their chest on this one and have not shared much information in regards to what is being added to the new version. Strange considering their name is “open”, so perhaps Microsoft’s AI event might have some more information.
Bookmark The Date March 16th 2023
Let’s face it could be hype and a complete coincidence. But it’s going to be worth a watch. The title of this news cast is “Reinventing Productivity The Future of Work With AI“. This will be held next Thursday March the 16th (8am PT).
It’s expected that the company will showcase the “future of AI”, including how it will work in productivity apps like Teams, Word, and Outlook, maybe using ChatGPT to write emails. We can expect to see more pushing of Microsoft products with an added feature, yeah “ChatGPT”.
So What Can GPT-4 Achieve?
Well this is where a lot of speculation comes into play. GPT-3.5 is already so much faster than previous models and even uses OpenAI’s Codex to provide users with examples of code, not to mention it’s the most advanced chat bot out there. So what’s next?
Well if it’s anything like the technology that Google and Meta have been showing off over the past few months. Then we have an inclining that they are going to be showing us Text to Video. It’s clearly the next step forward. Basically a user will enter a prompt (text description) and the AI model will generate a short video.
This is a huge advancement and has already been proven to be a bit of a challenge so it might be a big expectation to think that OpenAI has finally cracked it. But we can hope. Other suggestions are that the parameter will increase from 175 billion parameters to double to increase the response of the current chat bot. Speed is not really an issue as ChatGPT responds quicker than the human eye reads currently.
Or perhaps they are going to allow users to implement multiple models. So not only do you get a response from ChatGPT but also a video and maybe pictures? We can all just assume and speculate until the big day!
Let’s be honest it’s unclear what sort of parameters GPT-4 will have only OpenAI are aware of this, so anything you read online should be taken with a pinch of salt. Only recently OpenAI CEO Sam Altman shot down wild that suggested GPT-4 would increase the parameter count from GPT-3 from 175 billion to 100 trillion. Sure the AI space is increasing at break neck speed but not by that much.
OpenAI’s GPT-3.5 language model is already powering a range of chatbots, it’s well known ChatGPT tool, along with the new Bing Chat Microsoft (which is an OpenAI investor), Snapchat, Discord, and Slack.