Sparks are flying between the top players in the AI race.
Google and OpenAI’s executives have appeared to gently troll each other after the two companies had announced competing releases yesterday. On Tuesday, Google had launched its Google Gemini 2.5 Pro model, which outshone OpenAI’s o3-mini model on most benchmarks. Around the same time, OpenAI held a livestream announcing new image generation capabilities in ChatGPT.
At this point, Logan Kilpatrick, who is leads product at Google AI Studio, and has been the voice of Google on X for a while, responded to OpenAI’s announcement. OpenAI had shared a picture of a man wearing an OpenAI t-shirt writing on a blackboard the words “Livestream at 11am PT”. Image generation models have often had trouble with correctly rendering text, so the post was a gentle flex of OpenAI’s model capabilities.

But Google’s Logan Kilpatrick responded to the post. With a smiley face, he shared a picture of a person wearing a Google T shirt writing on a blackboard “Native image generation, available right now in Google AI studio”.

Kilpatrick appeared to be referring to how Google had launched its Gemini Flash Experimental model for image generation a few weeks ago, and had met with largely rave reviews. Kilpatrick seemed to be saying that Google already had the capabilities that OpenAI was about to release in its livestream.
But OpenAI had an ace up its sleeve. In the livestream, OpenAI not only announced capabilities to generate accurate text, but also demostrated abilities to transform pictures into different styles, such as those in the style of Japan’s Studio Ghibli, or create entire infographics with text and images. The Studio Ghibli images, in particular, went viral on X, with hundreds of users sharing the Studio Ghibli-style images they’d created.
As OpenAI CEO Sam Altman was presumably basking in the success of his announcement, he didn’t miss the change to rub it in to Kilpatrick. Yesterday, Kilpatrick had asked X’s users to share what they’d done with Gemini 2.5 Pro. “So what have you done with Gemini 2.5-Pro today? Please show impressive things only,” he’d posted. To this, Altman sarcastically replied: “my whole timeline is very impressive images! great ship”. Altman’s post was a dig at how OpenAI’s product had ended up going viral, as opposed to Google’s Gemini 2.5 Pro.

All of this is made even more interesting by the fact that Logan Kilpatrick used to work for OpenAI till March 2024, when he switched over and joined Google. But the competitive tension between Google and OpenAI is unmistakable at this point. OpenAI had been founded in 2015 partly to ensure that Google alone didn’t have a monopoly on AI. After launching ChatGPT in November 2022, OpenAI had managed to take AI mainstream. Google had been thought by many to be caught napping, and its products had generally seemed to lag OpenAI over the last couple of years. But in recent months, Google has appeared to hit its stride, and currently has the most powerful AI model in the world, ahead of OpenAI. OpenAI, though, is fighting back, and as Altman’s last post shows, is in no mood to easily give up its AI crown.