Sources: Google Brain and DeepMind have been forced to work together on a project known as Gemini to compete with OpenAI's GPT-4, after Bard's stumble
OpenAI's success in overtaking Google with an artificial intelligence-powered chatbot has achieved what seemed impossible in the past …
The InformationJon Victor
Context & Ripple Effects
Google’s response to Bard’s setback was not framed as another product tweak, but as a forced collaboration between two major research groups around Gemini. The significance is organizational: Google was concentrating previously separate AI capabilities behind a single GPT-4 competitor.
That effort subsequently moved toward early outside access to Gemini and a Bard model upgrade, while DeepMind described a design intended to combine LLM capabilities with AlphaGo-style systems. The story marks the internal starting point of that product and research arc.
First-order effects
Google Brain and DeepMind must align research priorities and execution around Gemini, reducing the separation between the teams for this project.
Bard becomes the immediate proving ground for Google’s renewed model effort after its public misstep.
OpenAI’s GPT-4 raises the urgency for Google to compete with a coordinated model-and-product response rather than separate lab initiatives.
Third-order effects
If sustained, the episode points to frontier-model competition pushing large platforms to consolidate research organizations around a smaller number of flagship models.
The longer-term differentiator may shift from maintaining distinct AI labs to connecting model research, deployment products, and developer access in one operating loop.
The trend: Frontier AI competition is driving big technology companies to consolidate fragmented research and product efforts behind flagship model platforms.
NEW: Prominent Google AI researcher resigned after warning Alphabet CEO Sundar Pichai and other senior execs that Bard—Google's rival to ChatGPT—was *using data from ChatGPT*. Big no-no in that world. https://www.theinformation.com/ ... w/ @jon_victor_ https://twitter.com/...
This @theinformation scoop is crazy: a top google ai engineer quit recently because Google was training its AI on OpenAI's ChatGPT. The models are copying each other! Makes an even bigger mush of authorship https://www.theinformation.com/ ... https://twitter.com/...
New: In a rare collaboration between DeepMind and Google Brain, software engineers at both Alphabet AI units are working together on a GPT-4 rival that would have up to 1 trillion parameters—a project known internally as Gemini. https://www.theinformation.com/ ...
Also these must be end times because the leaders of Google Brain and @DeepMind do NOT like working together but were forced to do so. https://www.theinformation.com/ ...
I'm old enough to remember Google derailing an industry conference on search by waiting till the morning of the event to accuse Bing of consuming data from a user installed browser toolbar that collected everything including the Google clickstream (as browser toolbars will do) ht…
ChatGPT learns off the web & our collective views/idiocy. Users document ChatGPT's wildest examples in ShareGPT. Google's Bard apparently learns from ShareGPT. At this rate, we can be sure that none of the pearls of wisdom we generated post Sept 2021 are used anywhere! ;-) https:…
I know people at DeepMind They spend more time worrying about ski trips than working I know people at OpenAI They are at the office until 3am multiple times a month I know people at Google They are more interested in their Tahoe ski lodges than grind Culture matters https://twitt…
A normal person reading this headline would wonder why Google has two giant AI teams that compete each other and they'd be right. https://www.theinformation.com/ ...