Best of Twitter week of February 24-March 2, 2025

Was Industrial Revolution possible in 300 AD? Lots missing: Didn't know how to make cheap quality steel. Didn't know how to harness a horse properly for agricultural activities. _Escape from Rome_ argues collapse of Empire was necessary institutional precondition for IR.

— Vlad Tarko 🌐 🏗️ 🚀 (@vladtarko)
4:21 AM • Feb 23, 2025

This company will literally store your backups on the moon. 😯 -

— Michael P. Frank 💻🔜♻️ (@MikePFrank)
5:28 PM • Feb 10, 2025

increasingly think that despite having had been a gainfully employed taxpayer for several years i may not have actually performed any work that has created real economic value

— met2llurgist (@met2llurgist)
2:43 AM • Feb 17, 2025

IVF mainstreaming is here, embryo selection is next. I maintain that I'm more accurate about stuff outside ML, even if it gores your sacred cows.

— Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞) (@teortaxesTex)
3:36 AM • Feb 19, 2025

I think @tylercowen is probably right that it may take years or decades for advanced ai/agi to show up in the productivity/gdp statistics

That's why I think it's generally better to focus on "when could r&d automation unlock strategically destabilising weapons technologies?"

— Haydn Belfield (@HaydnBelfield)
5:04 PM • Feb 23, 2025

I'm not at all joking about this. As others have pointed out the current culturally endorsed way to "do humanities" is far too slow and ossified to keep up. There are still academic papers that used GPT3.5 coming out that say LLMs can't do this or that

"Doing humanities things"… x.com/i/web/status/1…

— Katan'Hya (@KatanHya)
1:26 PM • Feb 21, 2025

What an amazing discovery from Scott Kennedy's lab! A new splicing system that recognizes transposons and removes them from mRNAs!

— Oded Rechavi (@OdedRechavi)
2:51 PM • Feb 18, 2025

Today I decided to try using o1 to assist with some math/CS research. Here's how it went.

For context: I'm currently developing an automated theorem-proving framework in Scheme, and attempting to produce formal proofs of correctness for some advanced numerical algorithms. (1/10)

— Jonathan Gorard (@getjonwithit)
3:30 AM • Feb 19, 2025

Halfway through university I was diagnosed with clinical depression. After a battery of tests and interviews with psychologists I eventually met with the psychiatrist who was to dispense my medication.

Instead he asked me a question that no one had ever asked.

"Why are you… x.com/i/web/status/1…

— Josh Whiton (@joshwhiton)
2:03 PM • Feb 17, 2025

I had a good rapport with a literature professor and talked to him one day about existential depression.  He just looked at me like I was crazy and said “you need a girlfriend, mate.”

— Natural General Intelligence (@realtimeai)
4:46 AM • Feb 18, 2025

Liang’s career proves how one man’s genius can shift global power dynamics. From TSMC to Samsung to SMIC, he has rewritten the rules of semiconductor warfare. 16/18

— William Huo (@wmhuo168)
10:30 PM • Feb 25, 2025

Anna's archive is doing a great job visualizing what parts of the human printed heritage are catalogued and preserved

— dreaming android (@pastaraspberry)
12:51 PM • Feb 26, 2025

i asked gpt4.5 what’s a deep truth that most people are not aware of
and then i kept asking

1. discomfort isn't just inevitable; it's essential. chase friction relentlessly.

— vittorio (@IterIntellectus)
11:25 AM • Feb 28, 2025