Google I/O 2024, the search giant’s flagship developer conference, will be held on May 14. Interestingly, it’s right on the day after OpenAI’s Spring Update, where the company is expected to announce a lot of exciting updates for developers and enterprises. And given the timing, I/O this year becomes an extremely important event for Google’s AI.
With almost every Google announcement over the past few years focussing on AI, this year’s event promises to be no different. Obviously, Gemini is likely to take centre stage, with Pixel phones also on the forefront.
Another area that Google is expected to put a lot of focus on is its Responsible AI deployment, given the recent debacle with its image generation platform Gemini creating historically inaccurate images. The company later apologised for “missing the mark” and CEO Sundar Pichai, in a recent interview, took full responsibility for it.
A lot of AI
There are reports of Google engaging in discussions with Apple to assist with generative AI features for iOS. At the same time, there are almost confirmed reports of OpenAI also partnering with Apple for similar features.
If the Apple deal materialises with OpenAI, which might happen today, or we may have to wait until Apple’s WWDC in June for further details, Google might have to devise another strategy to improve its AI game while continuing its partnership with Samsung.
Speaking of on-edge use cases, Google is also expected to announce Gemini Nano, its smallest AI model to compete with Microsoft’s Phi-3 or Meta’s Llama 3. The model has been in early access since its announcement in December 2023 and is running on Pixel 8 Pro, but it may be available for developers soon.
A developer preview of Android 15 was released last month, so Google might offer a deeper dive into its next mobile operating system, featuring deep Gemini integration. Additionally, we can expect previews of new versions of Google’s other platforms, including Wear OS, Android TV, and Android Auto, along with new AI developer tools for different OS.
One of the most important and loved features of Google’s generative AI has been on its Workspace platforms, such as Google Docs, Sheets, and others. The company has been constantly bumping up the capabilities of its products and might make further announcements about more capabilities on different platforms.
Since there are rumours that OpenAI might release its AI agents at the conference, Google is also expected to do the same. Recently, Google introduced Vertex AI Agent Builder, a platform that enables the easy creation of autonomous agents with little or no coding required.
A big focus on healthcare
Google has also found its moat in healthcare. In January, Google Research launched a new medical chatbot called AMIE, which specialises in expert-level differential diagnosis. Unlike the big tech’s previous AI models, Med-PaLM 2, which focuses on medical summaries or answering questions, AMIE serves as a diagnostic tool, generating differential diagnoses.
Meanwhile, Med-PaLM 2 is also expected to get a few updates.
Moreover in April, researchers from Google and DeepMind also developed Med-Gemini, a new family of highly capable multimodal AI models specialised for medicine. The model achieved an accuracy of 91.1%, surpassing the previous best by 4.6%. In multimodal tasks, the models outperformed GPT-4 by an average of 44.5%.
There is even a slight possibility of Google revealing an AI-developed drug at the event. Alphabet-backed Isomorphic Labs, along with DeepMind, recently released AlphaFold 3. The protein folding model predicts with 50% better accuracy.
In a recent interview, Demis Hassabis said, “Well, if you ask me the number one thing AI can do for humanity, it would be to solve hundreds of terrible diseases. I can’t imagine a better use case for AI. So that’s partly the motivation behind Isomorphic and AlphaFold and all the work we do in sciences.”
He believes that “revolutionising the drug discovery process to make it ten times faster” and more efficient and increasing the likelihood of passing clinical trials through better property prediction offers plenty of commercial value.
Hassabis said that Google DeepMind will combine the agent systems it developed for gaming with multimodal systems into large general models that plan and achieve goals.
“So systems that are able to not just answer questions for you, but actually plan and act in the world and solve goals… those are the things that will make these systems sort of the next level of usefulness in terms of being a useful everyday assistant,” he said. He further said that AI-designed drugs would probably be available in the ‘next couple of years’.
Something for developers
All these capabilities might make a big impact on Google’s integration of AI into its hardware devices at Google I/O. To up the ante with its competitors, Google’s biggest bet might also be to increase more AI models on its Google Cloud and Vertex AI platform.
According to Google’s page for the event, there are going to be several announcements about Google’s Cloud TPU and also accelerating its loads on PyTorch/XLA, which the company developed in collaboration with Meta.
Meanwhile, Google I/O 2024 promises several exciting updates and announcements for app developers. We might hear about a new version of Android Studio called Arctic Fox. This would include improvements in data security and privacy features.
New authentication methods such as face or fingerprint recognition for enhanced app security for Firebase are also expected to be announced.
Flutter, gaining popularity for cross-platform app development, might showcase a preview of Flutter 3 with support for new features in Android 15 and iOS 17, including better widgets and notifications.
When it comes to India, Google’s open-source model Gemma has been adored by developers for its Indic language capabilities. It is expected that Google may announce more updates to its AI models for building AI models in India.
All in all, it would be interesting to see how Google matches up with the announcements that OpenAI is going to make at its event.
