Genesis Aka ’s Post

Dive into the human-AI paradigm shift with Rachel Chalmers and Raiya Kind, PhD, on ep. 16 of Generationship. Explore the future of AI and consciousness. Listen now! https://lnkd.in/gn946HGM

"State of LLMs in 2024 : Reflections and Future Insights" - Webinar by Kern AI - Attended this interesting webinar (online) today. The exciting advancements shaping the AI industry, spotlighting the real-world use cases that transformed 2024, and the major trends set to redefine 2025 were presented.  The current landscape of open-source versus closed-source LLMs was reviewed, giving valuable insights to navigate this evolving field.

Can LLM's be steered to amplify creativity in their output? Today, during the 🔥Spotlight Talk🔥 at the NeurIPS 2024 Workshop on Creativity & Generative AI, Matthew Olson from our Cognitive AI team at Intel Labs described our research on how to use a mechanistic approach to steer LLM's in the direction of more creative output. Amazing work by co-authors: Neale Ratzlaff, Musashi Hinck, Shaoyen Tseng. Read our paper here, and stay tuned for more work on alignment and steerability of LLMs: https://lnkd.in/guZ3kr-c

Interesting seminar on how supercomputers can be used in life science research.

During Microsoft Research Forum: Episode 2, we heard from the experts on AI's revolution in scientific discovery and beyond. You can catch these discussions on demand now and learn more about science and technology research in the era of general AI. https://msft.it/6040ccyKK

For the past 6 months, I've been exploring the fascinating intersection of biology, medicine, and AI. Daphne Koller is one of my favorite female role models with her work in biology and AI. Following that thread, I came across a conversation with some of the best minds in AI at Davos this year. I loved the discussion on what's going to drive us towards unlocking the next level of discoveries in AI. Unsurprisingly, it will come from data - video and sensory data, especially in the intersection of biology and medicine where the level of information is immense. However, to unlock this level of innovation, Yann LeCun suggests that this intelligence will not be generative, but instead will need to plan and reason in a different abstract space with models of the world. I've always believed that general intelligence cannot be disembodied or in other words, without an ability to intervene in the world and run its own experiments. As Daphne highlighted this further, it is necessary for AI to understand the causal effects of the world. I highly recommend it: https://lnkd.in/gsmi3KpY #inspire #aiinnovation #womeninscience

We reported the discovery of a novel structural class of antibiotics at Integrated Biosciences, Inc. a few months ago in Nature. Now, this work has been highlighted in a report from the AI Working Group of the President's Council of Advisors on Science and Technology (PCAST). "By leveraging AI algorithms to navigate the vast chemical space of potential molecular structures, researchers were able to identify a novel compound with potent antibacterial activity against a wide range of pathogens, including those resistant to existing antibiotics. This type of discovery-based method shows great promise for providing new insights about molecular structures and functions, for example leading to whole new classes of antibiotics."

This recorded conversation from Obvious Ventures discusses generative science as the next big wave in AI innovation. While most people are already familiar with the revolutionary capabilities of large language models and generative AI (which is trained on the Internet), generative science is about building large science models (trained on the physical sciences, like biology, chemistry, geology, and physics). As the Obvious team says, "This is the most world-positive application of artificial intelligence, one that can really move humanity forward." Watch the full recording here: https://lnkd.in/g7HBPKad

Gen AI is expected to have a profound impact on life sciences - from accelerating the discovery of new #therapies to enhancing the #patientexperience in #clinicaltrials to improving the success rate for trials. Learn more in this new research from Everest Group and Medidata Solutions. #artificialintelligence #clinicalresearch #Medidata

Take a read through this white paper to see how clinical trial stakeholders expect to utilize Gen AI 💡