OpenAI has announced GPT-4o, showcasing a new conversational AI model that can take prompts from multiple forms of media simultaneously. But just an hour before the announcement was underway, Google posted a teaser for its own Gemini, showing off basically the same conversational capabilities.
Of course, a single teaser isn’t going to reveal as much as an full announcement. But based on what was shown, it looks like this version of Google Gemini can do just about everything that OpenAI’s GPT-4o can. The teaser shows the AI model looking – with a phone’s camera as its eyes – at the Google I/O stage being set up ahead of the dev conference, which starts tomorrow, on 15 May. It goes on to describe what it sees, and mentions hoping that something AI-related will be announced.
One more day until #GoogleIO! We’re feeling 🤩. See you tomorrow for the latest news about AI, Search and more. pic.twitter.com/QiS1G8GBf9
— Google (@Google) May 13, 2024
On one hand, the timing of the announcements and this teaser make sense, if you believe in the first-mover advantage and considering the similarities of the products. But on the other, Google has mostly locked the more impressive bits of Gemini behind a paywall. It remains to be seen if what is being teased here will be available to the free Gemini Pro, or locked behind the paywalled Gemini Ultra. In comparison, OpenAI has announced that GPT-4o will be gradually rolled out to free users, even if it comes with usage limits.
Of course, these questions will probably be answered at Google I/O tomorrow, as well as a clearer description of the new capabilities being made available to Gemini. This will also probably when Android 15 gets its official unveiling.
(Source: Google / X)
