Digital Decluttering
summary
Gartner 2030 claim that synthetic data will eclipse real data in AI models anchors the talk and the concrete takeaways are tools, numbers and results: demos of Rust crates (microbi rust crate, rig, fake_it, synthetic-data-sh, lace), a conversion benchmark (10,000 GenBank files: 25 minutes vs 4 hours Rust vs Python single‑thread), trained protein sequence generators (spike model outperforms flu, model ~1.7× smaller than data), Elixir ML recommendations for metrics, and cautions about privacy–utility tradeoffs and edge‑case coverage.
AI-written summary · about
chapters (19)
- 0:00Digital decluttering intro
- 1:52What is synthetic data
- 5:04Unstructured data scale
- 7:16Synthetic data in biosciences
- 9:47COVID medical imaging case
- 11:15DNA as big data
- 14:58DNA for data storage
- 15:19Engineered biological sequences
- 17:27Protein structure frameworks
- 19:36Rust synthetic data tools
- 20:18Rust vs Python in bio
- 23:13Microbi-rust for genomics
- 26:34Evaluating synthetic data
- 28:02Protein generation with Rust
- 32:18Metrics for synthetic data
- 33:01Digital twins in medicine
- 35:57Gen AI for dataset storage
- 37:45Future of synthetic data
- 38:28Q&A
related talks
36mArchitecting Data: A Deep Dive Into the world of Synthetic Data
30m🤯 No data? No problem! Synthetic data to the rescue
17mFocus on Data, on the Road to AI
20mSynthetic Data for Privacy, AI Fairness and Democratizing Data Access
20mSynthetic Data: The new Super Power for AI
25mThe Importance of Synthetic Data in Data-Centric AI
12mGenesis Final Round: Betterdata
42mRise of the Machines: Achieving Data Security and Analytics with AI