Google DeepMind at ICML 2024

Exploring AGI, the challenges of scaling and the way forward for multimodal generative AI

Subsequent week the unreal intelligence (AI) group will come collectively for the 2024 International Conference on Machine Learning (ICML). Working from July 21-27 in Vienna, Austria, the convention is a world platform for showcasing the most recent advances, exchanging concepts and shaping the way forward for AI analysis.

This 12 months, groups from throughout Google DeepMind will current greater than 80 analysis papers. At our sales space, we’ll additionally showcase our multimodal on-device mannequin, Gemini Nano, our new household of AI fashions for schooling known as LearnLM and we’ll demo TacticAI, an AI assistant that may assist with soccer techniques.

Right here we introduce a few of our oral, highlight and poster displays:

Defining the trail to AGI

What’s synthetic common intelligence (AGI)? The phrase describes an AI system that’s a minimum of as succesful as a human at most duties. As AI fashions proceed to advance, defining what AGI might seem like in apply will turn into more and more vital.

We’ll current a framework for classifying the capabilities and behaviors of AGI models. Relying on their efficiency, generality and autonomy, our paper categorizes programs starting from non-AI calculators to rising AI fashions and different novel applied sciences.

We’ll additionally present that open-endedness is critical to building generalized AI that goes past human capabilities. Whereas many current AI advances had been pushed by current Web-scale knowledge, open-ended programs can generate new discoveries that reach human data.

Scaling AI programs effectively and responsibly

Creating bigger, extra succesful AI fashions requires extra environment friendly coaching strategies, nearer alignment with human preferences and higher privateness safeguards.

We’ll present how utilizing classification instead of regression techniques makes it simpler to scale deep reinforcement studying programs and obtain state-of-the-art efficiency throughout totally different domains. Moreover, we suggest a novel strategy that predicts the distribution of consequences of a reinforcement learning agent’s actions, serving to quickly consider new situations.

Our researchers current an alignment-maintaining approach that reduces the necessity for human oversight, and a new approach to fine-tuning large language models (LLMs), primarily based on recreation concept, higher aligns a LLM’s output with human preferences.

We critique the approach of training models on public data and only fine-tuning with “differentially private” training, and argue this strategy might not provide the privateness or utility that’s typically claimed it does.

New approaches in generative AI and multimodality

Generative AI applied sciences and multimodal capabilities are increasing the artistic prospects of digital media.

We’ll current VideoPoet, which makes use of an LLM to generate state-of-the-art video and audio from multimodal inputs together with photographs, textual content, audio and different video.

And share Genie (generative interactive environments), which may generate a variety of playable environments for coaching AI brokers, primarily based on textual content prompts, photographs, images, or sketches.

Lastly, we introduce MagicLens, a novel picture retrieval system that makes use of textual content directions to retrieve photographs with richer relations past visible similarity.

Supporting the AI group

We’re proud to sponsor ICML and foster a various group in AI and machine studying by supporting initiatives led by Disability in AI,
Queer in AI,
LatinX in AI and
Women in Machine Learning.

When you’re on the convention, go to the Google DeepMind and Google Analysis cubicles to satisfy our groups, see dwell demos and discover out extra about our analysis.