Deep learning and neural networks are shifting how we approach many problems, making neural networks the best solution for a growing number of challenges.

The new technology of large language models is creating unprecedented opportunities for startups, enabling tasks that previously required extensive research teams to be accomplished more easily.

Our group investigates the hypothesis that large amounts of compute can solve interesting problems using neural networks, which has led to significant advancements in unsupervised learning at scale.

The same neural network architectures that excel in computer vision also apply to other fields, such as speech recognition, demonstrating the interconnectedness of AI tasks.

In using neural networks for problems like computer tomography, we are exploring how to combine handcrafted models with neural networks, particularly in how we feed them prior information.

The last two years have been dominated by large AI models, which have proven to be popular and effective, but we are now entering an era focused on complex workflows around these models.

Machine learning emphasizes the task of making machines learn from little initial knowledge to improve over time, with deep learning techniques utilizing neural networks being particularly effective in recent years.

The breakthrough with large language models (LLMs) has accelerated the development of high-quality AI, which is now advancing faster than holographic AR technology.

The current low-hanging fruit in AI is using large language models to perform knowledge work and integrating them into familiar user interfaces without changing user behavior.