Google Joins the Open Source Game with Gemma

Welcome Renaissance Creator,

The arc of technology is long, but it bends towards decentralization.

Open source and permissionless innovation are two of the most powerful ideas of the internet era, and will serve as the bedrock of the new Renaissance.

Google’s announcement of Gemma, a suite of open-source LLMs now available to the public, is an important step toward that end.

Let’s dive in.

The One Big Thing

Open Source is becoming the new norm in AI

We are heading to a world where “OpenAI” is the only AI company left without a single open-source large language model.

For the rest of the industry (Meta, Google, Mistral, Anthropic) open source is becoming the norm, and it is a beautiful thing to see. While the models that these companies release are generally less performant than proprietary ones like Gemini 1.5 or GPT-4, they are generally small enough to run on a laptop, and are miles ahead of anything the industry had even a year ago.

For all of the concerns about AI centralization and ensuing consolidation of resources and power, it seems so far that things are trending in the opposite direction.

In this early stage of AI development, companies creating models have to fight over a resource even more scarce than chips: capable developers. Right now, the fight is for researchers and developers who can build cool applications on top of the core infrastructure that Google and others are providing. To do that, they have to attract the best talent to their platform and ecosystem.

There is no better way to do so than to create a blank canvas that developers can play on and build with. While most developers will still opt for Gemini 1.5 or GPT-4 for production applications, the fact that Google is releasing these models, their weights, and the inference code means that developers will be able to build on top of these foundational models themselves and tweak them to their liking.

The developers that do so will almost certainly be using Google’s Gemini to build applications, along with related tools like Google Workspace.

The broader impact of this move is that having an open source software (OSS) model may become table stakes for any company looking to operate seriously in the space.

This means that in the very near future, anyone in the world will have free access to a wide range of high-performance models they can run on a basic laptop. When you combine this with the fact that models are only getting more powerful, you are going to get some crazy outcomes.

Some that we are already starting to see…

Artists’ Gallery

Open source models are letting people create text to video and the results are beautiful

Get ready to start having two way conversations with your pets… I wonder how many pet breakups this leads to…

Tools

Must have tools for every Renaissance creator to add to their toolkit:

Deep Tech

The newest and coolest in the research world that you need to know about:

Closing Thought

Work With Us!

The AI Renaissance is coming and we are building the best community of the people making it happen.

Contact us to sponsor your product or brand and reach the exact audience for your needs across our newsletter and podcast network.