OpenAI took the wraps off its latest Generative Artificial Intelligence (AI) model, known as ChatGPT-4o, at an event earlier this week. The upgraded AI chatbot can now hold conversations in real-time and interact across text, pictures and video — identifying and solving a maths equation from a photo in one demo.
ChatGPT-4o will be available to all users in the coming weeks, thanks to the reduced cost of running this model for OpenAI. Paid subscribers will still unlock a number of perks, including the ability to generate images from a written prompt via Dall•E 3 within the ChatGPT interface.
Microsoft has invested billions of dollars in OpenAI and deployed ChatGPT across its suite of productivity software, allowing it to take an early lead in the generative AI race.
Microsoft Copilot, which costs £19 per month to unlock all featuresin Microsoft Word, Excel, and Powerpoint, is powered by its GPT model. The talkative AI assistant also makes an appearance in Microsoft Edge — its answer to Google Chrome, as well as Windows 10 and Windows 11 desktop operating systems.
But despite this unprecedented amount of support and record $10 billion investment, OpenAI announced that its first dedicated desktop app for ChatGPT-4o would not appear on Windows 10 or Windows 11.
Instead, it’s reserved for macOS.
The desktop ChatGPT application is rolling out now, with paid ChatGPT Plus subscribers first in the queue to get their hands on the software. “We will make it more broadly available in the coming weeks,” OpenAI has promised in a blog post about the launch.
Until now, ChatGPT has only been accessible from a web browser or via the Android, iPhone, and iPad app. With the arrival of the macOS version, desktop and laptop users will be able to leverage the AI chatbot separate from their web browsers. There should be no difference in speed between any of these ChatGPT apps, since all of your requests are handled by the cloud and OpenAI doesn’t use on-device processing.
However, there are a number of unique features available to those who install the ChatGPT app on their MacBook or iMac. First up, you’ll be able to instantly ask ChatGPT a question from anywhere in the operating system with the keyboard shortcut Option + Space. This works like an alternative to Spotlight (Cmd + Space) and Siri, which are both preinstalled as part of macOS.
Mac users can also take and discuss screenshots directly in the app from the Share menu in macOS. You’ll need to be running MacOS 14 and Apple Silicon (M1 or better) to use the incoming app.
Installing the dedicated ChatGPT app on desktop has a number of advantages, including the ability to summon the AI assistant from anywhere and send screenshots for it to analyse
OPENAI PRESS OFFICE
OpenAI has pledged to launch a native Windows app for ChatGPT “later this year”. It’s unclear whether it will have the same functionality enjoyed by Mac users.
It’s unclear exactly why OpenAI has chosen to support macOS ahead of Windows, despite the latter being used by considerably more devices worldwide. In a statement to Axois, a spokesperson for OpenAI said it started with the Mac because “that’s where more of its users are.”
If the browser-based version of ChatGPT receives more visitors from Apple’s desktop operating system than Windows, it does make sense to prioritise the development of an app for the Mac App Store.
The widespread roll-out of Microsoft Copilot means those who use a Windows PC can already tap into the functionality of ChatGPT from their laptop or desktop — without the need for a dedicated app from OpenAI.
“The decision could also be influenced by the fact that more people at OpenAI use macOS than Windows.” Microsoft-focused blog Windows Latest posits as another possible explanation for the choice.
Microsoft is working on its next major update to Windows 11, which will introduce a number of new AI features across the operating system. Those familiar with the functionality describe the AI capabilities as “groundbreaking”.
Whispers from inside the Redmond firm suggest Windows 11 users can rely on natural language when searching for files, applications and contacts across the operating system. So, instead of searching for the title of a Word document, PC owners will be able to search for a document that “was created last week and has a chart about Q4 revenue” or “PDF Matt sent on WhatsApp on Monday”.
Microsoft is also looking to supercharge Live Caption, so that its AI can translate several different languages in real-time during a video call, video, or phone call. Samsung recently added real-time translation for text messages and voice calls with its Galaxy S24 smartphone range.
Microsoft is widely believed to be training a new, in-house AI language model large enough to compete with Google Gemini and OpenAI, an exclusive report from The Information revealed earlier this month. This would reduce the company’s reliance on the start-up behind ChatGPT.
The rival model, referred to as MAI-1 within the company, is overseen by recently hired Mustafa Suleyman, the Google DeepMind co-founder and former CEO of AI startup Inflection, two Microsoft employees with knowledge of the effort confirmed to the publication. The exact purpose of the model has not been determined yet and will depend on how well it performs.
LATEST DEVELOPMENTS
However, MAI-1 is believed to be “far larger” than the previous smaller, open-source models Microsoft had previously trained which means it will be more expensive, according to the report.
Microsoft could preview the new model as soon as its Build developer conference later this month, the report said. Microsoft declined to comment when contacted by writers from Reuters.