Transformers - Opportunities and Limitations in Real World Applications

[In-person/hybrid event]:
Here is the link if you want to join online!

Five months ago, Microsoft and NVIDIA introduced the largest and the most powerful transformer-based language model Megatron-Turing Natural Language Generation model (MT-NLG). How excited should we be about this mega-model trend?

Join the talk with Karl Moritz Hermann, Founder and CEO of Saiga, to discuss this topic in more detail:

• Have super large-scale language models changed how we should think about AI in the context of natural language processing?
• What are the limitations of hyper-scale models, particularly when applied to real-world applications?
• Can hyper-scale models compensate for a lack of multi-modal or embodied learning?
• And, of course: are large scale models conscious?

About Karl Moritz Hermann
Karl Moritz was the founder and CEO of AI research startup Dark Blue Labs in Oxford which sold to Google in 2014. Afterwards, he worked as a researcher at DeepMind in London and later in Berlin with over 8,000 citations, 50 publications, and 6 patents. Karl Moritz also has some years at McKinsey Berlin. He gained his PhD degree in Machine Learning & Natural Language Processing from Oxford.