And here we are just focusing on Google and OpenAI. There have been other major announcements, particularly in AI video synthesis and AI research papers, that have been similarly overwhelming. Some very important releases may get overshadowed by the rush of news from major tech companies and the distraction of the holiday season.
Willison weighs in on the AI frenzy
“I can’t remember this much of a flood of December releases from anyone, I’d expect people to be quieting down a bit for the holidays,” said independent AI researcher Simon Willison in a conversation with Ars Technica. He also gave a rundown of this busy month in a post on his own AI blog, hardly knowing how to make sense of it all.
“I had big plans for December: for one thing, I was hoping to get to an actual RC of Datasette 1.0, in preparation for a full release in January,” Willison wrote. “Instead, I’ve found myself distracted by a constant barrage of new LLM releases.”
Willison sees this part of this flurry of activity as a side-effect of the heated rivalry between OpenAI and Google. In the past, it’s been common for OpenAI to surprise-release a product to undercut an expected Google announcement, but now the shoe is on the other foot.
“We did see a rare example of Google undermining an OpenAI release with Gemini Flash 2.0 showcasing streaming images and video a day before OpenAI added that to ChatGPT,” said Willison. “It used to be OpenAI would frequently undermine Google by shipping something impressive on a Gemini release day.”
The rapid-fire releases extend beyond just these two companies. Meta released its Llama 3.3 70B-Instruct model on December 6, which Willison notes can now run “GPT-4 class” performance on consumer hardware. Amazon joined the fray on December 4 with its Nova family of multi-modal models, priced to compete with Google’s Gemini 1.5 series.
