Real progress is being able to make and characterize 10,000 materials from sufficiently complex composition space in a day. Really doesn't matter that much how far computation is scaled beyond SoTA

2024 AI/ML Publication Trends in Science: 📈 Strong growth across domains continues! Whether you're a graduate student, developer, researcher in industry, professor, postdoc, or citizen scientist - these numbers reflect one aspect of our collective progress. Let's look at the

I think we should start writing proposals with months 'level granularity

Good thing: my evening entertainment now includes watching Despicable Me and other cool movies Not so good thing: I am asked to dance

🪄 Welcome to the era of quantum sheep-puting! Now the data centers can become truly green powered, ecological, and directly translating ML to the real world in the form of products. I wonder how DI Vincenzo criteria are satisfied for quantum sheep-puters? linkedin.com/posts/sergei-k…

When your friend is Kyle

Ben Blaiszik 1. Taylor Sparks 2. Janosh 3. Maxim Ziatdinov 4. Sergei Kalinin 5.Andrej Karpathy (i know lofty request but would be cool) 6. Tian Xie 7. Olexandr Isayev 🇺🇦🇺🇸 8. Tim Duignan 9. Adeesh Kolluru 10. Brandon Wood 11. Dr. Kamal Choudhary ...

Interestingly, the ML academic twitter starts to come back in bits and pieces after 2 years of draught

AI comedy night - exploring career pivot to stand up comedy youtube.com/watch?v=jNP76Y…

You know your civilization messed up big time when the hardest part of doing anything is getting permission to do it. We should be fighting nature, physical law, entropy, etc, not government rules and bureaucracy.

In AI4Science, we need AI for decision making and physics for science

Like mother like daughter

First episode is out!🎊 I spoke with Sergei Kalinin (UT Knoxville,Pacific Northwest National Laboratory) on how AI is transforming microscopy, from imaging to autonomous experimentation. Stay tuned for the second episode next week! Listen here: open.substack.com/pub/ml4sci/p/s…

The positive side effect of ML is increased astronomy knowledge: Egyptians -> Ptolemeian model -> Copernikan -> Kepler laws -> Newton laws -> (general relativity)

How to make double edged feedback to a student: "your response to reviewers is super good. Now you train all new group members to do it".

Any advice on good books on classical design of experiments?

tickle.utk.edu/mse/solar-cell…

Is MatMiner still a thing?

After listening to Jingle Bells 20 times, I convinced the baby that Du Hast by Rammstein is German Jingle Bells. Is it:

There is no computation without representation, and optimization without reward function