AI shifted the level of abstraction we can work on. For future generations, manual coding will probably feel like very low-level programming (maybe like assembly to most of us). I wonder how people will approach computer science in the future.

Jupyter notebooks are great for data science + engineering… but I’ve always wanted something more flexible.

If you’d like more info, feel free to reach out.
#dbt #duckdb #datascience #dataengineering #software

Cosine Similarity and Euclidean Distance produce the same ranking over normalized embeddings in vector retrieval systems. Mathematical proof is a fun exercise.

We are experimenting with new interfaces for scientific literature in the age of AI. In our first version, we use AI to transform abstracts into a standardized format (objective, method, outcome). This improves skimming through papers with less cognitive load.

Let us know what you think.

🧬 Our roadmap paper on how #QuantumComputing could shape the future of #SingleCell analysis is now out in @NatuRevMolCell!

We explore the intersection of physics and biology to redefine precision medicine. ⚛️🔬

Read it here: www.nature.com/articles/s41...
🧵👇

#NatureReviews