Hello from Berlin, where earlier today we hosted the first of three I/O Connect events happening this year. These events are some of my favorites because we get to bring our people and technology to you – the developers who choose to build with Google.
This year, we were joined by over 1300 passionate developers, from over 100 countries in Europe, the Middle East and Africa. And the energy was palpable – from the keynote stage to the demo floor, office hours, technical sessions, panels, and more - it’s days like these that make me feel so grateful to be a part of the Google Developer community.
EMEA is home to 16 million developers, almost 40% of the world's developer talent, which meant there was a whole lot of ground to cover! On stage, Tim, Wieland, and I highlighted Google’s progress in AI (of course), including a fireside chat with Germany’s Christina Decker, Departmental Head of Digital and Innovation Policy, Federal Ministry for Economy and Climate in Germany, on responsible AI development.
Here are some highlights from what we announced both on stage in Berlin and elsewhere in the world in celebration of this event!
The AI ecosystem is moving fast, and we’re committed to continuing to innovate and make AI accessible and helpful to every developer. Here are some of the most recent updates:
Today, we’re officially releasing Gemma 2, in both 9 billion (9B) and 27 billion (27B) parameter sizes, to researchers and developers globally. Gemma 2 features improvements in performance along with significant built-in safety advancements. Get hands-on with the newly launched Gemma 2 without hardware requirements in Google AI Studio, or you can also download Gemma 2’s model weights from Kaggle and Hugging Face Models, with Vertex AI Model Garden coming soon.
We’re making it even easier and useful to build with the Gemini API with updates like expanded access to the 2 million context window in 1.5 Pro for all developers, context caching in both Gemini 1.5 Pro and 1.5 Flash, and code execution capabilities in the Gemini API. For those starting out, you can pre-register for a free online course that we’ve built in partnership with Udacity to teach developers how to build with the Gemini API. [Blog]
Join thousands of developers from all over the world who have signed up for the Gemini API Developer Competition for a chance to win a custom electric 1981 DeLorean! To spark your imagination, we unveiled an interactive I/O Crossword demo, built on our multi-platform tools, and open-sourced the code. We hope you’ll use it as inspiration for your own submission to the competition, open until August 12, 2024.
We were excited to celebrate the active contributions to many of the form factors powered by Android, including wearables and cars from our engineering offices in Nairobi, London and Munich, as well as our progress toward an AI-assisted web.
Developers in EMEA can now work with Gemini to generate code with just one client, access helpful resources, and even troubleshoot errors all with Gemini in Android Studio.
With Gemini Nano integration in Chrome we are making it easier to deliver powerful AI features to Chrome’s billions of users without having to worry about capacity, cost, fine tuning or prompt engineering with high level APIs, like translate, transcribe, or caption. Today, developers can sign up to explore further use cases with the Prompt API, an experimental API that lets you send natural language instructions to an instance of Gemini Nano in Chrome — all on device and in the browser.
Starting today, developers in Europe can now take advantage of AI-powered insights within the Chrome DevTools Console. Gemini will provide explanations and solutions to DevTools errors and warnings, significantly streamlining your debugging process. This is all brought to you by a team based in Munich.
Developers in EMEA can use this Chrome DevTools extension to understand the impact of third-party cookie deprecation on their site, and plan site migrations for a smoother transition.
We highlighted investments in low latency Cloud computing, like the first ever subsea fiber-optic cable directly connecting Africa with Australia, which we called “Umoja” or “ unity” in Swahili. We also talked about how we’re investing through Cloud to make it easier for developers in EMEA to transition to compliance with local regulations with updates like:
Unlock your full potential for digital transformation with Google Cloud solutions, designed to help organizations, especially those in regulated industries or the public sector. These solutions provide developers with control over their data residency, encryption methods, and key management, regardless of their industry. We also announced expanding its sovereign cloud offerings to Spain, Italy, and Saudi Arabia in addition to existing partnerships in Germany and France.
Build scalable solutions everywhere, with new data centers across the region. Last year, we launched new Cloud regions in Doha, right here in Berlin-Brandenburg, and Damman, and this year, we announced the Johannesburg cloud region in South Africa, which benefits from the high-capacity fiber optic cables under land and sea.
Our accelerator programs have supported nearly 350 startups in 54 countries across EMEA. 40% of these companies are women-led. We deeply believe in the effectiveness of the Google Developer community, so we announced:
Developers in EMEA focused on net-zero emission efforts can apply to the Climate Change Accelerator and EMEA-based entrepreneurs building AI-first solutions can apply to one of two AI-First Europe accelerators for a chance to leverage the best of our technologies, expertise, and network to scale their impact.
Our mission continues to be to make AI accessible and helpful to every developer. We're just getting started on this exciting journey to make it easy for you to bring Google’s developer tools together more easily for your own projects.
Our next stop is Bengaluru in a few weeks, where we have more exciting news to share about exactly that.
See you soon - in the meantime, keep on building!
オンデバイス カスタム LLM 用の AI Edge Torch Generative API
Gemini 1.5 Pro 2M context window, code execution capabilities, and Gemma 2 are available today
Model Explorer: Simplifying ML models for Edge devices
Fine-tuning Gemma 2 with Keras - and an update from Hugging Face