@Puͣkiͧte̍.com 🇱🇻

30K posts

@Puͣkiͧte̍.com 🇱🇻 banner
@Puͣkiͧte̍.com 🇱🇻

@Puͣkiͧte̍.com 🇱🇻

@WHUT

whut's happening? → Earth Sciences 🌏 Mathematical GeoEnergy (Wiley/AGU, 2019) 🌊 https://t.co/IGGOHc2YC1 https://t.co/v9iiFNoYzq

heart of @ilhan -- MN 5th ! Se unió Mart 2008
1.5K Siguiendo718 Seguidores
AD Tippet
AD Tippet@BelAves·
@Jake_W Science is done by scientists, people with their own biases. Many of the scientists in the 1970s wanted Erlich to be right and ignored any doubts. Julian Simon did not and who was right?
English
4
0
31
1.9K
Caleb Wachter
Caleb Wachter@CalebWachter·
The whole point of science is to be wrong, and to prove statements incorrect. It's literally how science operates. But people being people try to use science to validate or confirm things. It cannot do this. It will never do this. It is fundamentally incapable of ever doing this.
English
6
1
12
1.3K
Marian L Tupy
Marian L Tupy@Marian_L_Tupy·
"65 million Americans would starve to death" - Nope. Obesity is a bigger problem. “England will not exist in the year 2000” - Nope. Still here. nytimes.com/2026/03/15/boo…
English
25
65
415
18.8K
Marty Milligan
Marty Milligan@martyamilligan·
@charlesmurray Redefining the scientific method as peer review was a bold move. Let's see how it works out...oh God no.....
English
1
0
8
304
Friends of Narcissus
Friends of Narcissus@cultofnarcissus·
@charlesmurray Peer review has become a consensus of mediocrity protecting itself from the threat of ingenuity.
English
1
0
19
336
@Puͣkiͧte̍.com 🇱🇻
@3RenChengHu @Noahpinion It will steer to the consensus quite heavily, criticizing the parts of the paper that are genuine insights. Have to go through several iterations of arguing with an LLM for the insights to "take", and they still may not.
English
1
0
0
6
Against Narrative
Against Narrative@3RenChengHu·
@Noahpinion Why would you not use AI for peer review? You can discard bad suggestions, and it might pick up something you missed.
English
1
0
2
51
Noah Smith 🐇🇺🇸🇺🇦🇹🇼
Honestly, my sense is that >50% of academics always just faked peer review. A bunch would just skim the paper and misread it and demand nonsensical changes ("Reviewer 2"), while others would just say "Why don't you cite MY paper" ("Reviewer 3"). At least AI will read the paper!
Natalie Khalil@natalienkhalil

Headlines like this are alarming because we don't have a standardized way to quantify AI performance for peer review. We've spent the last three months working on a benchmark. We're going to make it open source and preprint it!

English
22
24
315
38K
Prof. Ryan Katz-Rosene
Prof. Ryan Katz-Rosene@ryankatzrosene·
Actually Clack it’s just basic fucking science. The natural world balanced emissions with sequestration over century-long timescales; humans then created a one-way pump transferring Carbon from the lithosphere to the atmosphere.
Prof. Ryan Katz-Rosene tweet media
Peter Clack@PeterDClack

To blame humans for 100% of the carbon dioxide rise is simplistic and misleading. It suggests the natural world - with its tectonic shifts, mid-ocean ridge volcanism, feedback mechanisms and massive water vapor flux - all suddenly vanished in 1850. The greening of the planet (confirmed by NASA) proves the biosphere is actively utilising this CO2, yet the crisis narrative treats the Earth as a passive victim rather than a self-regulating, dynamic system. It conveniently tries to frame humans as careless or even as evil. It is more like a case of self loathing. You are completely ignoring the natural world and the real heavy lifters of Earth’s climate: the oceans and the hydrological cycle. The focus on a 150-ppm rise in CO2 bypasses the reality that the oceans contain 50 times more carbon than the atmosphere. Oceans hold 1,000 times the heat capacity. They contain 91-92% of all the entire world's retained heat energy. Water is 1,000 times denser and heavier than the air, a fine inert vapour by comparison We are currently observing the tail end of a multi-century adjustment, as the deep ocean responds to solar and orbital cycles that began long before the industrial revolution. This is how the natural world turns, very slowly. The CO2 rise isn't a human footprint crushing the natural world out of existence, it’s an echo from a complex feedback loop involving the entire planetary engine.

English
54
122
1.1K
22.3K
Paul A. Szypula 🇺🇸
Paul A. Szypula 🇺🇸@Bubblebathgirl·
President Trump on the Strait of Hormuz and America’s energy independence: “We don’t need oil. We have all the oil we need for ourselves.” This is exactly why it was wise for Trump to ramp up domestic oil production. Independence makes America great.
English
41
211
1.5K
13.2K
Gen
Gen@ccgencc·
@getjonwithit Verbosity is part of what makes an artefact worthy of our time. Its like ornaments on the facade of a building. Without it its lifeless. Also llm are much more verbous than humans and in a very unappealing way. Companies just hide the thinking tokens
English
1
0
0
90
Jonathan Gorard
Jonathan Gorard@getjonwithit·
I think one of the conclusions we should draw from the tremendous success of LLMs is how much of human knowledge and society exists at very low levels of Kolmogorov complexity. We are entering an era where the minimal representation of a human cultural artifact... (1/12)
English
187
489
4.5K
741.1K
Ben Schulz
Ben Schulz@schulzb589·
@ramez Stack enough sigmoids and you hit exponential improvements. Moore's Law wasn't just Dennard scaling.
English
1
0
1
78
Ramez Naam
Ramez Naam@ramez·
This is the core question in AI recursive self-improvement to me. If AI RSI is mostly about writing code, or running experiments with tweaks over current algorithms, we shouldn't expect a takeoff to ASI. But if AI models can truly find breakthroughs in machine learning algos and paradigms, then anything is possible. Unfortunately, there's no real way to quantify the likelihood of the latter. And given current AI capabilities I don't see a lot of evidence that they're poised to discover breakthroughs that are truly "out of distribution" of current knowledge.
Haider.@slow_developer

Sam Altman says another breakthrough beyond transformers could be coming, and models are now smart enough to help find it On products, AI creates a huge chance to rebuild entire product categories and make new things possible "but AGI will look like just a warm-up for what comes next"

English
6
5
25
8.6K
Bud D. from Benoit
Bud D. from Benoit@soloflexerus·
@FareedZakaria @CNN @ZelenskyyUa The 'oil shock' is not America's problem. With our own reserves, and that of Venezuela, we Americans have plenty of oil. In fact, we make good money selling our excess production on the world market; this is good for America.
English
1
0
0
21
Fareed Zakaria
Fareed Zakaria@FareedZakaria·
On GPS, at 10am & 1pm ET Sunday on @CNN: I’ll be joined by Ukrainian Pres. @ZelenskyyUa for an exclusive interview … • The same Iranian-designed drones being fired across the Middle East have been used by Russia against Ukraine for years. I’ll ask Pres. @ZelenskyyUa about the help he is offering now—and where his own country’s war is heading • What can we expect from Iran under new leadership? I’ll talk w/ CNN Global Affairs Analyst @ksadjadpour & Lisa Anderson, dean emerita of @ColumbiaSIPA • The Iran war has triggered one of the worst oil shocks in decades. I’ll examine the implications w/ @JasonBordoff, who served in the Obama admin’s Nat’l Security Council as sr. dir. for energy & climate change
English
7
8
62
13.1K
Dudes Posting Their W’s
Dudes Posting Their W’s@DudespostingWs·
This is how the most dangerous job you never hear about keeps your world running every single day. Roughnecks are the frontline crew of oil and gas drilling operations.
English
1.2K
3.5K
28.6K
2.6M
Yujia Zheng
Yujia Zheng@YujiaZheng9·
@WHUT @YiMaTweets @ylecun Yeah, and for machines they need to identify nonlinear systems. All neural networks are with nonlinear activation layers, and it is quite hard to find a real world process that is purely linear.
English
1
0
1
88
Yi Ma
Yi Ma@YiMaTweets·
In system theory, it is called "linearization"... which has been studied and used for decades. Honestly, folks, there is no need to invent or introduce any new terminology. Remember, there is rarely anything new under the sun...
Ying Wang@yingwww_

What is a good latent space for world modeling and planning? 🤔 Inspired by the perceptual straightening hypothesis in human vision, we introduce temporal straightening to improve representation learning for latent planning. 📑: agenticlearning.ai/temporal-strai…

English
22
52
736
112.9K
Chennakesava Kadapa
Chennakesava Kadapa@chenna1985·
ML in the last few decades has been successful primarily because it either renamed or reinvented stuff just for publications. CS folks never seem to bother with learning anything established in the core fields of maths and engineering.
Yi Ma@YiMaTweets

In system theory, it is called "linearization"... which has been studied and used for decades. Honestly, folks, there is no need to invent or introduce any new terminology. Remember, there is rarely anything new under the sun...

English
2
0
22
1.4K
Honest Alder (not dishonest crank Alder)
Yes you are a bad person @latimeralder as you constantly lie and deceive. You debate like a 7 year old. You are like a snake oil salesmen slithering away from scrutiny. You ignore data you don’t like or pretend it doesn’t exist. One of the most dishonest people on the planet.
Honest Alder (not dishonest crank Alder) tweet media
English
3
3
21
207